DE10107792A1 - Contact management between service providers and service users or searchers the invention relating particularly to locating required services via the Internet, with weighted search criteria used to optimize the search process - Google Patents

Abstract

System for contact management between service or goods providers (A), searchers (S) and intermediaries (V) using a communication network, especially the Internet. Accordingly data records are selected from a database evaluated and transmitted to the searcher. The selection can also be made on the basis of service or work time and location. A number of evaluation criteria are used and weighed according to importance so that an overall search value index is obtained by which the search results are sorted. The search list is then transmitted to the searcher. Based on the usefulness of the matching list the evaluation criteria can be modified.

Description

The invention relates to a system for contact mediation between provider A and seeker S via Intermediate V using communication network plants. The invention is particularly applicable to Providers and offers according to the goals of Searchers select and sort and the contact between these providers and offers and the To manufacture seekers.

Contacts between seekers S and Manufacturers A manufactured in three different ways: with the help of printed directories, search engines on the Internet or web catalogs on the Internet.

Printed directories have been used for searching and Selection of objects and providers proven. you will be only updated annually. The sorting of the entries is fixed and usually takes place according to industry and provider Name. Increase with each ad and its size Chances of success and fixed costs of the providers, therefore choose appropriate industries and concise descriptions.

On the Internet, the searcher gives information via PC or telephone his search data, with a search engine or a Web catalog. They are looking for suitable ones worldwide Objects, select them as hits, create the sorted hit list and transmit it to the searcher a short version of the website. After request  Texts, images, links to appear on the website further pages or the possibility to email Send. Possibly. displayed phone numbers are coming tarfeld, no link to this telephone connection. To Editing a page allows the searcher to hit return list, request other pages or the End search. Its cost is low.

Providers design and change their pages themselves and determine the search terms that are free for them. You pay for the design of the pages and the used computer capacity and mostly get invoice which contains the number of clicks or emails, Telephone contacts but not recorded.

With search engines, the searcher gives free, logical linkable search terms. The search engine checks whether the websites you know contain the terms hold, select the pages as hits and sorts them, e.g. B. according to machine text analysis. There Selection and sorting determine the chance of success, providers design their websites in accordance with the Selection and sorting algorithms of important searches machines and not just factual. Abuse, e.g. B. Word spamming is only partially used by search engines recognized.

Therefore, some search engines analyze behavior from stakeholders to evaluate their website to recognize. One approach is to analyze links of those who are assumed to be neutral. Some providers however, establish link communities and are not new tral. A second approach is sorting by number Clicks or total contact times. This makes decisions of the seekers' evaluations, but quanti tative sizes as qualitative sizes. So points z. B. 1 contact from 30 seconds to 1 satisfied Seekers there, 30 contacts with 1 second each but sooner to 30 dissatisfied.  

If a provider chooses an appropriate and an inappropriate matching search term, then the unrelated apt term the selection for the applicable term didn't grip. However, the incorrect term improves the sorting station, even if the term is incorrect a few clicks and the number of clicks increases. The provider therefore optimizes the sorting station and its Chance of success at the expense of the seeker.

In the case of web catalogs, the searcher chooses an object class out. Further entries can be freely selected or by an input mask tailored to the object class Right. Then within the web catalog for This object class searched for permitted data sets.

The admission is personal and results in one Ranking number by which the hits are sorted. This also allows images and sounds to be checked and checked Exclude manipulation through word spamming. The disadvantage is the human effort. He gets ground additionally not charged to the provider, but must in Urgent cases are paid. It also seems problematic that a person has the worldwide relevance of an internet side judged.

Market researchers use the internet for marketing and Data transfer. The market as an instrument research stands in the way that only part of the household te has internet connection and the use very is different. There are often concerns about the Data protection, especially with personal data. If a sample is representative can therefore only sel be determined. It can be assumed that the Members of the sample also within the subset the internet users are not representative.

The user can be asked explicitly if he is the Researching data agrees. Disadvantages of the survey are the time expenditure of the user and therefore more limited  Application. Analyzes of the user association avoid polling situation and time expenditure, they are possible without the consent of the user, e.g. B. through cookies stored on the user's PC or Log files that show the course of use. On Another advantage is that all users are recorded and so the representativity is assured.

Using the example of the yellow pages, the US. Sector telephone directory www.smartpages.com and the German company directory www.yellowmaps.com should some important disadvantages of traditional directories being represented.

Yellow pages are based on data from Deutsche Telekom AG. Their conventional form is regional books Validity and sorting by industry and provider name. They contain for all business customers of Germans Telekom AG a small, not billed Display. The provider must provide further or larger advertisements pay, they result in e.g. B. for Berlin approx. 2,000 advertisements sides with a circulation of 1.305 million and thus 2,610 million pages for the 2000/2001 edition.

For seekers, the ads in an industry are one Hit list for this search term. The list shows immediately all hits, their approximate number and every hit completely. It is not very clear. The providers try by size and design of the Ads to gain the interest of the searcher and point in part to performance time, e.g. B. "opened by , , , to . , . "or" 24-hour service ", or Place of performance, e.g. B. "all-district", "delivery service" or all branches.

The publishers of the yellow pages put on the Internet at www.gelbe-seiten.de a nationwide one Search mask ready in which the provider name that Object class via branch, a search term for brand /  Products / service, and the place to be entered can. The location can be entered via Place name, zip code or area code at which Advanced search also by street or house number. For Berlin also gives the option of taking www.gelbe-seiten-berlin.de and there the place to be entered via street and zip code or city District. Search results in hit lists with page 15 each Providers with name, zip code, telephone and possibly a symbol. The provider name is a link to more Information, the symbol shows the type of this Information.

In the Internet variant, the representation of the Locations clear, since the search is regional is limited and data is shown in several stages become. The data are mainly from the publisher updated.

Service location and time are not input fields. Will the Performance needed at unusual times or locations then starting from the hit list, each ad must be opened to see if there is any information gives. The location can have up to five parameters can be entered, but only considered Business addresses. Delivery service is ignored, Searching for providers nearby is not possible.

When searching with location information, a company in the Usually found when his business address is in the area defined by the location information. A Exceptions are so-called "out-of-town recommendations" "where a provider extends its offer Assign areas. Another exception are Branch operations that are not recorded in many cities and then not in searches with location information being found.  

The US called. Local telephone directory and the German company directory are similar to the yellow ones Pages. They use the same search fields, phone Numbers are only comment fields, place of performance and -Time is not a search condition. Unlike the Yellow pages are the provider data from the provider, the receives access authorization from the intermediary alisiert. It also searches for distance by distance Airline possible, regardless of performance.

The update by providers reduces the effort of Intermediary and enables faster updating and complex inputs. For the environment search the automatically entered by the seeker Assigned location code and then with the location code of Address of a selected provider, the distance determined. The selected providers will increase after sorted by distance.

While addresses and location codes of all providers are recorded the searcher does not need his address enter. Then when determining the distance uses an address in the middle of the search area. The environment search is carried out for all services on the Basis of the business addresses and without limit for the Distance. It makes sense if the performance matches the Business addresses is bound, e.g. B. Visit to the swimming pool. It is e.g. B. not useful when repairing a Washing machine or remote diagnosis of a fax machine.

With some intermediaries (e.g. www.goto.com) you can Providers a payment offer for each click on their Submit website. The provider chooses the amount of his bid and pays to the intermediary who Internet pages per property class according to the decreasing amount sorted and shows all visitors the bids. The Providers can bid for future clicks at any time  change and bid for "ecommerce" in late September 2000 up to $ 5.99 per click.

With his bid, the provider improves the sorting space, with rising costs. He's at his optimum if the bid and benefit are the same per click. The The amount of the benefit is determined by the probability possibility of an exchange of services with seekers and the average amount of the resulting benefit. The bidder will therefore be the higher, depending greater the proportion of seekers who Service sharing agree, and the average Volume of this exchange are.

Because also the advantage of the searchers per click through this Sizes is determined, the height is more conditional Provider disadvantages for the descending sorting of the Results. As a disadvantage, payment amounts are due to the related earnings also from the perspective of Mediator well suited.

A special switching system provides taxi They represent the offers of taxi companies, the public concessions and Have regulations on place and price: there are a compulsory driving area, within whose orders too specified prices must be carried out. The Price is determined by the route traveled by the passenger determines and does not depend on the approach route.

Over 95% of all trips are as immediate as possible and from Location of the seeker to begin with. Because place and time the selection of a provider at the start of the journey determine, they are here as a place of performance and Performance time used. Service area of a taxi is - regardless of the business address of the sub taker - always the immediate surroundings of their respective location that changes over time.  

Passengers do not want individual provider numbers search, call and ask if a vehicle is in their Environment is operational. Solve this problem taxi Central offices that act as mediators here. You bundle the locally limited and changing offers of many Providers at any time the mandatory driving area covering offer.

Taxi centers arrange for staff to be accepted Orders with the help of a specially developed EDP System. The street and house number of the Address and name to be reached when making pre-orders additionally the performance time, in special cases also Search data, e.g. B. "Passenger with dog", or comments, z. B. "Waiting in the restaurant".

The IT system arranges the street and house number of the address to be approached a zone and each zone one of Day of the week and time dependent buffer time. There are the order for award free, with immediate orders after Entry, for pre-orders at around the buffer time before the performance time.

The providers report to the EDP system by radio that they are ready to order. Your location becomes machine recognized and reported to the IT system. It knows that operational providers in the zone of performance locations and their waiting times, checks compliance with the Search conditions and mediates the order to the Provider with the longest waiting time. Is in the zone no provider itself, then the IT system checks the Neighboring zones.

The first provider reads the place of performance in a display and time and comment, almost always takes the job reports this to the EDP system and drives to Of performance. If he does not accept the order, then sets the computer system waits for 0 and offers the Order to the second and possibly other providers.  

Taxi centers are mediators with personal mediators for only one object class in a limited Territory, with personal reporting of the performance Readiness of the provider and machine detection of their changing performance area. The providers have identical prices and are from the perspective of the seeker homogeneous. The selection criterion is almost always the one by Zones divided into classes by distance between Seeker and provider, sorting criterion is the Waiting period. Failure to accept an order results in Provider the disadvantage that in the following searches the Sorting place is worse because of the set to 0 Waiting period.

So far, phone numbers are just a comment field and not a link. The medium is changed for calls and the telephone number entered personally. Disadvantage are the effort for typing and the bad evaluation availability of telephone contact.

Existing industry solutions are specialized and who made in small numbers. The pro rata Hardware and software development costs are high, in contrast to Internet-based solutions.

The invention is therefore based on the object To create system which has the disadvantages of fixes existing solutions and especially firstly ensures an improved selection of objects, second, the selected objects according to interests of the seeker and after this evaluation the Third, the selected objects are sorted provides automatic contact making and finally by evaluating the processes Optimization of the contact mediation realized.  

This object is achieved by the Features in the characterizing part of claims 1 and 2 in cooperation with the features in the generic term. Advantageous embodiments of the invention are in the Subclaims included.

A particular advantage of the invention is that amounts of data stored in at least one database

• - be evaluated in such a way that at least one in the evaluation criterion value- Number and / or quantity number and / or weight number determined and this to a total value number be compressed or if a quantity of data only has a value number, this value number as a total Value number is adopted,
• - selected according to the specifications of an assessor B, in a hit sorted by the total value number list and the contact between Assessing B and at least one selected Amount of data or provider A of an amount of data is automatically created by immediate selection or after transmission of at least part of the Hit list for assessors B according to individual duel decision is realized.

Another advantageous embodiment of the system according to the invention consists in the fact that amounts of data stored in at least one database

• - be evaluated in such a way that at least one in the evaluation criterion value- Number and / or quantity number and / or weight number determined and this to a total value number be compressed or if a quantity of data only has a value number, this value number as a total Value number is adopted,  
• - selected according to the specifications of an assessor B, in one sorted by the total value number Hit list can be recorded, and the contact between Assessing B and at least one selected Amount of data or provider A of an amount of data through automatic immediate dialing or transmission at least part of the selected data sets is produced at the evaluators B.

The invention is based on at least partially shown in the figures and with the help execution examples explained by table representations play will be described in more detail.

It shows:

Fig. 1a, b representation of the system components and the flow of information and data between the evaluators and these components.

The combination of the information and data flows between the two parts a and b of the figure is indicated by the numbers 1 to 8 ;

Table 1 Example of a possible data assignment with input masks for provider master data,

Table 2 Example of a possible data assignment with a (general) start mask and the Entry mask for object data of the object class "taxi",

Table 3 Example of a possible data assignment an input mask for performance data, with Special features of the object class "taxi",

Table 4 Example of a possible data assignment a (general) start mask and the on Input mask for search data of the object class "Taxi",  

Table 5 Representation of the evaluation of search data and actual values for selection and evaluation and linking to provider bids and Ratings by searchers for sorting of data volumes using the example of the object class "taxi",

Table 6 Presentation of individual evaluations using Analysis of the processing of the hit list by an evaluator (B) using the example of Object class "taxi",

Table 7a Calculation of the new evaluation of a Criterion (passenger satisfaction) from previous evaluations of the criterion and new individual evaluation,

Table 7b Definition of the evaluation of a criterion (Passenger satisfaction) by means of given start values when recording a new property (here: taxi),

Table 8 Object and performance data of a first Provider of copy services at Registration,

Table 9 Object and performance data of a second provider of copy services upon registration,

Table 10 Object and performance data of the second Provider of copy services to first change of data,

Table 11 Object and performance data of the second Provider of copy services to second change of data,  

Table 12a, b shows the sort order and Contact mediation as a result of Search data and object and performance data (see Tables 8 to 11).

The invention relates to the mediation of contacts between providers A and seekers S. These strive for Exchange of services and need contacts to seekers or providers of services. The Mediator supports providers and seekers operates a system for arranging contacts.

The system uses communication networks and data banks. There can be contacts between providers, intermediaries and mediate seekers even over long distances, connect multiple databases and complex Save large amounts of data quickly to edit.

A data set can be distributed across several databases. The intermediary V operates databases himself, which are referred to as internal memory (cf. FIG. 1), and accesses databases called external memories (cf. FIG. 1) operated by third parties if these contain the data he needs. It stores the ID number and the external locations of the data volume and the data that are not available to it externally in the internal memory.

The data stored or used by the intermediary consist of the amount of data and other data, e.g. B. Master data of evaluators B or algorithms. she expediently with a hardware and Existing electronic data processing programs processing system processed.  

A data set is evaluated by a Rating B using a rating method M for a Evaluation criterion. It always contains a value judgment as a qualitative component called value number W and can be a quantitative, called quantity number M. Component included. Also number of weights or weight The time of the assessment are possible components of the assessment.

The evaluation criteria are divided into groups, according to rating B and rating method M, and with BMi designated. I is the consecutive number of the Criteria within a BM group.

If there are several evaluations for one evaluation criterion BMi conditions, then the value numbers become this Ratings, if necessary taking into account the quantity numbers and the evaluation times, the number of values WBMi and, if applicable the quantity number MBMi and the time tBMi of the BMi criterion.

If no assessment was made for a BMi criterion, then the quantity number MBMi is 0, otherwise greater than 0.

From the total of all data amounts within the Database will be the amount of data that the specifications of the evaluator B, searched and selected.

This is done for each amount of data in the database is included, checked whether they meet all mandatory requirements Fulfills. If within the review of a data set it is determined that a mandatory requirement is not is fulfilled, the test is terminated and the Amount of data excluded. A data set is called Hit selected if it meets all mandatory requirements Fulfills. A mandatory requirement z. B. the Set of two specifications linked by "or" his.  

After checking all in the database contained data is for each hit from the Value numbers WBMi of the evaluated criteria, if necessary under Consideration of the weight numbers GBMi and the quantity Numbers MBMi, the total value number and, if applicable, the Weight number and the quantity number of the data quantity calculated. In the case of criteria without evaluation, the Weight number GBMi set to 0.

Then the hits are sorted according to their total value W sorted in descending order, so that the best ten rated hits are at the beginning of the hit list. Hits with the same total number of values can e.g. B. al be sorted alphabetically. But they are also lagging behind Introduce current sorts according to other regulations bar.

The switching system then makes the contact between assessor B and at least one Amount of data or its provider A.

In a first variant, called immediate selection the evaluator immediately with the one at the beginning of the Hit list of available data or their provider A connected. After this contact, the evaluator becomes involved connected to the next hit in the hit list because the evaluator ends the mediation.

The intermediary V uses the immediate choice if the machine-sorted hit list probably from Evaluating is taken over. If this is not the case, then the second variant is used.

The evaluator B is at least a part of the am Amount of data at the beginning of the hit list telt. The assessor then decides whether and with which of the transmitted data sets the contact deepened. He can end all existing contacts Request contacts for further hits or the ver or end the connection to the operator.  

In both variants, the contact of the evaluator can are limited to the amount of data, part of the Amount of data or the provider or with at least one Part of the amount of data and the provider. In Exceptional cases, e.g. B. with special requirements or Cancellation of the mediation, there is no contact.

The stored data amounts contain object data called data of an object. Possible objects are arbitrary things, living things, also people, companies and immaterial, like information. Especially at The object becomes information with the amount of data at least partially transmitted.

The object data are supplemented by master data of the Provider, which are referred to as provider data. she are expediently only saved once per provider and the object via the identification number of the offer ters assigned. For services that the provider provided themselves, object and provider are partially correct match. This applies in particular to personal Services.

From this master data and derived quantities, z. B. Environment of the provider's business address, too are the performance data that differ Performance descriptive data. They are crucial for the assessment of the provider A.

Reviews of property, provider or performance data can also be part of the data set. she become part of the amount of data before and part inside half of the current selection.

Property, provider or performance data or ratings can also be used as search data to select data sets be used. For selection according to ratings minimum values of value numbers.  

From the viewpoint of the seeker S in particular, a desired performance by handing it over to him. place and time of delivery are therefore considered as a place of performance and Performance time. You are next to performance price of this service important performance data.

Checking the performance data is complex. In the Selection is therefore preferably first object and Provider data checked, then the place of performance and finally performance time and performance price based on due to transport time and costs from the place of performance can depend.

Provider A often lists multiple service locations as considered equivalent. Therefore, the intermediary V several data fields for power zones OAL ready in which of the providers he accepts Enter individually, in subsets or as a total can. One of these data fields is as a performance zone OAL1 reserved for business addresses or current ones Locations with the possibility of transferring services.

The place of performance OSL that the seeker S is aiming for he into a data field provided by the intermediary V. Enter OSL. Places here are geographical locations. They can be characterized by their complete postal address or part of it Location codes, region or phone number with country, Area code and / or connection.

The searcher S can restrict the check as to whether a quantity of data fulfills the performance location OSL to the performance zone OAL1, e.g. B. if he wants to be present when the service is provided. There are two different types of tests:

• 1. A Is the OSL location within an OAL performance zone?
  Here it is checked for all locations in all performance zones OAL to be evaluated whether the location matches the location OSL of the seeker S. These locations are selected, the performance zones containing them are marked.
• 2. B Is the place of performance OSL within the maximum distance LSX1 determined by the seeker S?
  With this variant, the distance to the performance location OSL is determined for all locations in all performance zones OAL to be evaluated and it is checked whether it meets the maximum distance LSX1. The location with the shortest distance is selected from these locations. The performance zone containing it is marked.
  The distance from locations can be determined by the distance as the crow flies, the length of the shortest route or the time or costs expected for the transport.

All data sets without at least one marked Performance zones are excluded.

The agent V Data fields ready, in which the provider A the of him accepted time period DAL or combinations of Order time tAA and service time tAL can enter.

Seeker S can also do what he is looking for Time period DSL or combinations of order time tSA and performance time tSL in data fields that the Intermediary V provides.

Without considering the duration of transports, the intermediary V uses the equations

tAL = tAA + DAL

DSL = tSL - tSA.

It provides for the consideration of the transportation time DAT data fields ready, the performance zones OAL assigned. The provider A can in this Data fields that he usually needs Transport duration and in another data field Duration of the DAE, which he needs without transport, enter.

The mediator V calculates according to these values

DAL = DAT + DAE.

He can enter the transport duration DAT and Simplify the duration of the DAE and provide the provider A further data fields assigned to the performance zones OAL provide AAT ~ v and AAT ~ n and the path duration dAT ~ v and dAT ~ n before and after Provision of the service and the duration of performance DAE ~ 1 for the first and DAE ~ 2 for additional units of the Power.

The mediator V provides the searcher S with a data field ASL for the number of desired power units and uses these values to calculate the transport duration DAT and the service duration DAE

DAT = DAT ~ v + DAT ~ n

DAT ~ v = dAT ~ v × AAT ~ v

DAT ~ n = dAT ~ n × AAT ~ n.

DAE = DAE ~ 1 + (ASL - 1) × DAE ~ 2

The standby times of provider A can be eliminated can be viewed by occupying data fields for Start tAB ~ ya and end tAB ~ ye of his readiness periods AB ~ y.

The mediator V uses this data to calculate the standby time DAB ~ y and its sum DAB

tAE ~ a = tSA + DAT ~ v

tAE ~ e = tAL - DAT ~ n

DAB ~ y = tAB ~ ye - tAB ~ ya

with tAB ~ ya <tAE ~ e and tAB ~ ye <tAE ~ a

DAB = sum y (DAB ~ y) + tAE ~ e - tAE ~ a + tAB ~ ya (y ~ min) - tAB ~ ye (y ~ max)

with tAB ~ ya <tAE ~ e and tAB ~ ye <tAE ~ a.
and checks compliance with the performance time tSL with one of the following variants A and B:

• 1. A standby times are disregarded: planning from the beginning, starting from the placing of the order
  The intermediary V sets the times tAA of the order receipt at the provider A and tSA of the order placement by the seeker S in accordance with
  tAA = tSA,
  calculates the service time tAL of the provider and checks with the equation
  tAL <= tSL,
  whether the time tAL fulfills the searcher's performance time tSL.
• 2. B Standby times are observed: Planning from the end, starting with the handover of the service
  The mediator V equates the service times tAL of the provider A and tSL of the seeker S in accordance with
  tAL = tSL,
  calculates the times tAE ~ v of the earliest start and tAE ~ n of the latest end for the provision of the service, calculates the available standby time DAB and checks with
  DAB <= DAE,
  whether it is sufficient for the performance.

The agent V provides data for the service price fields ready, which he the performance zones OAL and the Performance time tAL assigned and that with that of the provider A accepted best from amount and currency unit the performance price PAL can be described.

The seeker S can achieve the desired performance price PSL with amount and currency unit in data field the one that the intermediary V provides.

The mediator V enables the differentiation of Price of a service unit. It provides data fields ready for that from provider A and seeker S accepted prices PAL ~ 1 and PSL ~ 1 the first Power unit and the prices PAL ~ 2 and PSL ~ 2 other power units.

The mediator V calculates according to the equations

PAL = PAL ~ 1 + (ASL - 1) × PAL ~ 2

PSL = PSL ~ 1 + (ASL - 1) × PSL ~ 2

and then checks for different currencies, taking exchange rates into account

PAL = <PSL,

whether the amount of data exceeds the searcher's upper price limit S Fulfills.

The selection of data sets is carried out according to specifications B. The intermediary V provides for this, Search data called specifications called search fields Data fields ready. It saves the search data at least temporarily in the databases and relieved so further searches with changed search data.

The mediator V enables input for each search field SXk of search limit LSXk, search condition BSXk and search target ZSXk. It leaves more than a combination of Search limit and search condition and more than one search target to.  

The combinations of search limit and search condition determine the selection or exclusion of data sets You are necessary for an active search field. The Search goals are values aimed at by the evaluator B. she affect sorting, not selection optional and can also be search limit.

When selecting, the intermediary V checks whether a Amount of data meets the specifications of all search fields. To he determines the associated one for each search field SXk Actual value ISXk of the amount of data and checks whether this actual Value all combinations of search limit LSXk and Search condition BSXk of the search field met.

The intermediary V aims for low costs and high ones Topicality. He avoids human resources, gives enter a few basic data for each evaluator, awards Rating B passwords and permissions for input and modification of, hereinafter referred to as data summarized amounts of data, further data and Algorithms.

The mediator entitles provider A to part of their Provider data and the object data of your objects enter and maintain and own data sets as Evaluate providers. He leaves seekers S that Use the switching system anonymously and authorize it after registration, save data and algorithms, z. B. Individual weight numbers GBM ~ Should be for groups of evaluators B and evaluation methods M or Machine evaluation rules. He also enables you to check inputs through test runs and then correct it.

Entering a rating B can be done by preferably action controlled by his will respectively. It can also be done by omission an action or by refraining from an action  within a period of time that is usually a Third pretends to assessor B.

The mediator V analyzes all of them possible object data and defines object classes that can be one or more stages. The object classes enable data sets based on the data contained in them Classify object data. The object classes can can also be used as a search field.

The switching system enables classification, but does not force them. The intermediary V can Define the "General" object class and its amount of data assign without a special object class.

A uniform structure of all one object class assigned data volumes simplifies the entire process processing the amount of data, especially the input, Modification, management, selection and evaluation. This The intermediary V already provides the structure when it is created new amounts of data ready. It allows that per amount of data only part of the structure is activated and that the activated part is different, both for one Amount of data over time as well as at a time for several data sets of an object class. He updated when the structure of a Object class the structure of already saved Amounts of data.

Uniform evaluation criteria for each object class result in comparable evaluations of the associated Amounts of data and increase the quality of the hit list. If an evaluation criterion by its first Rating is activated for all selected Data sets of an object class are either uniform is active or not active, then everyone gets it Amounts of data have the same weight number.  

The mediator V provides the evaluator B with input masks ready to enter, edit and Facilitate data assessment.

In the simplest case there is only one input mask with only one data field. So z. B. in the search without Object classes called Unstructured Search at least one in a freely writable field freely selectable search term, possibly with logical Shortcuts to be entered.

When searching with object classes that are structured Search is called, the fields correspond to Entry of data volumes and search data. This improves selection and sorting.

A sensible design of the first input mask is a freely writable field for unstructured searches and a pull-down menu with object classes for Structured search to use both of these Link data fields so that only one field is described and as a third field a switch for to use the input of further data. If the Evaluator B does not want to enter any further data, then the search for a data field if necessary using pre-assignments. If on the other hand, if he wants to make further entries, then offered him additional input masks.

Different types of data fields have specific ones Advantages: Freely writable fields result in a high Flexibility, e.g. B. for data of the object class "General", for one criterion, e.g. B. Highest speed of a vehicle, writable fields reduce the input effort and show that Assessing B which data is important. graphic editable fields facilitate the input of more complex Issues, e.g. B. unlimited performance zones by means of a pointing device and one who evaluates it  transmitted map. Pull-down menus can be multiple Contain levels, show the available variants and ensure that only permissible variants are entered become. Sliders allow quick adjustment of parameters and switches entering bipolar Sizes, e.g. B. Yes / No.

The matching of the input mask to the respective con text improves their application. In case of processing processing or searching for data sets an object class is determined, then after entering the object class a matching input mask is provided. The input mask can also show the current process, e.g. B. Entry, search or evaluation, and the result achieved Consider processing status or part of one more tiered tree structure.

In the standard case, pre-assignments reduce the Input effort and, in a special case, enable exactly coordinated inputs. You improve the use of the Systems for beginners and for experienced users.

Use and use of pre-assignments can Location and time of performance are illustrated. Your pre-assignment takes into account that seekers S the desired performance often at their current location OS0 and want to receive as soon as possible.

The mediator V recognizes the current location of the Seekers OS0 and / or the provider OA0 automatically over an existing communication network, e.g. B.

• - for stationary connections via dial-in nodes or Connection,
• - give mobile connections with cells active Transmitting and receiving systems to the cell,
• - generally by bearing by means of transmission and reception installations of a wireless network or
• - via signals from the Global Positions System  and occupies the OSL location with OS0, so that usually the seeker does not have to enter any places.

The mediator V assigns a standard period of time DBL to each object class, which is accepted by almost all providers A and seekers S for the provision of services of this object class. At the start of a search, he pre-assigns the order times tSA and tAA with the current time tS0 and, after entering an object class by the searcher, calculates S using the equation

tSL = tSA + DBL

the pre-assignment of the service time. Normally the searcher therefore does not need to enter times. He can also see how long DBL is customary in the market.

When determining the usual DBL service period the mediator can add more besides the object class Consider data, e.g. B. the current time tS0 current location OS0 of the seeker, the place of performance OSL and the number of ASL of the desired performance units.

Weight figures GBM ~ Should registered seeker S enable the searcher to pre-assign the weights Replace GBM with individual values and permanently to be stored in the internal memory. These pre-assignments the searcher can search for the current Adapt situation.

The intermediary V also uses pre-assignments in order to To use knowledge of the assessors B. He leaves changes in individual cases, these changes are automatically evaluated from, e.g. B. using data mining, and fits using Control loops on its default settings. Possible Applications are object classes and hit lists.  

The agent can, for example, B. determine how often providers A and seekers S used them. He can summarize rarely used object classes and Break down commonly used object classes. Also Input for the "General" object class Notes for new object classes.

The hit lists contain the hits in a through the mediator's sorting algorithm Sequence. The evaluators B decide which one Hits they want to deepen the contact, and assign so on possible optimizations of the sorting algorithm out.

Provider A can evaluate the amount of data Seekers S and the intermediary V, which therefore as Rating B can be called. In addition to these at the Mediation Participants are other evaluators, z. B. Expert G, approved. The assessors B can individual natural or legal persons or formally or informally related groups of people be and act directly or indirectly.

A provider A is always the provider of the respective Amount of data and can offer more or all in the Datasets of data stored. If one Providers of other amounts of data evaluate this amount of data, he is not a provider, but e.g. B. Seeker S or Expert G.

The searchers S search for data sets with the The main goal is to find suitable amounts of data for them. The Assessment of data volumes can be a secondary goal.

The mediator V operates the mediation system. He can simultaneously offer one, several or all Amount of data, one of several searchers or the one umpteen seekers.  

The experts G have the main goal of increasing the amount of data rate. They are neutral and not providers at the same time a quantity of data you have assessed. Your commitment he possible, complex objects or amounts of data to be constantly checked and evaluated. Are experts z. B. the operator of web catalogs, trade magazine independent experts or the foundation Warentest. In the case of appraisers, the objects are practical try it out, then consider their rating alongside Expectations also know-how.

An evaluation can be made according to different evaluation methods M done, e.g. B. Personnel assessment, assessment according to regulation and decision-oriented Rating. The assessor B can do several at the same time Evaluate amounts of data or different criteria.

In the case of personnel rating P, a certain one is already Existing data volume assessed by personnel. The value judgment is at the discretion of the evaluator B and will explicitly given, e.g. B. by specifying the ranking number an Internet homepage by the operator of a web Catalog or by pressing a button for e.g. B., Very good 'with value number 1,' medium 'with value number 0.5 or 'Bad' with value number 0.

In the, preferably machine evaluation according to The assessment of data volumes is carried out according to the regulation a rule that an assessor B previously and regardless of this amount of data. The Regulation can be stored in the database and dated Rating B can be changed at any time. The change can take effect immediately or be restricted, e.g. B. on not yet existing, not yet after this Regulation assessed or not yet selected Amounts of data.  

Possible components of the regulation are its validity area, e.g. B. an object class, the input variables, z. B. object data or provider data or value numbers based on previous assessments, the baseline, z. B. value numbers, and the algorithm for the output quantities as a function of the input quantities. examples for this evaluation method are evaluations of the HTML Codes or the meaning of all the links of an internet Page in accordance with the requirements of intermediary V.

With decision-oriented assessment E become Hit a current selection, preferably machine nell, rated. The intermediary V analyzes the dispatch the data from and to the evaluator B. He recognizes the underlying decisions, the implicit Valuations are, and transforms them into value numbers and quantity numbers. This method requires at Assessing B no additional effort. she can automated and used extensively, e.g. B. at the determination of the click numbers per website.

The actual value of a data set ISXk and the search data a search field SXk determine the selection of the data quantity. They also affect the sorting of the Tref fer, if the assessor B at least in the search field enters a search target ZSXk. The search target and Search limit coincide, e.g. B. when choosing one Company car, whose engine power according to the company is limited by an employee who has a strives for the highest possible engine power.

For search fields with a search target, during the search, namely after selecting a quantity of data and before calculating their total value number W, one for the search field SXk Value number WBVi calculated that only applies to this search.

This calculation can be done in two steps:  

1. Calculation of the degree of target achievement zSXk independently from the evaluator B

The degree of target achievement zSXk is calculated from the actual value ISXk, search limit LSXk and search target ZSXk according to the general formula

zSXk = zSXk (ISXk, LSXk, ZSXk)

with 0 = <zSXk = <1
and zSXk = 1 for ISXk = ZSXk
and zSXk = 0 for ISXk = LSXk and LSXk << ZSXk,
where the degree of target achievement zSXk is normalized to values between 0 and 1, is always equal to 1 for actual values that are a search target ZSXk, and is always equal to 0 for actual values that are a search limit LSXk and not at the same time a search target ZSXk , The actual values can first be transformed so that all actual values ISXk are in the range between 0 and + infinity.

The mediator V allows more than one search target, e.g. B. lower and upper search target ZSXk ~ u and ZSXk ~ o. It also allows multiple combinations of search condition and search limit, e.g. B. with lower and upper search limit LSXk ~ u and LSXk ~ o. If the upper and lower search target do not have the same value, namely ZSXk, then target achievement level 1 applies to all actual values between the two search targets

zSXk = 1 for ZSXk ~ u = <ISXk = <ZSXk ~ o,

For actual values between a search target ZSXk and a search limit LSXk, which is not also a search target ZSXk, linear functions are preferred, e.g. B.

zSXk = (ISXk - LSXk ~ u) / (ZSXk ~ u - LSXk ~ u)

for LSXk ~ u = <ISXk = <ZSXk ~ u

zSXk = (ISXk - LSXk ~ o) / (ZSXk ~ o - LSXk ~ o)

for ZSXk ~ o = <ISXk = <LSXk ~ o.  

If the evaluator B permits values of any size for the actual value ISXk or strives for the highest possible value, then the value + infinity applies for the upper search limit LSXk ~ o or for the upper search target ZSXk ~ o. Linear functions are not suitable in these cases and are replaced, e.g. B. by exponential functions with the constant Ck or by reciprocal function according to one of the formulas

zSXk = e EXP - (ISXk / Ck) for 0 = ZSXk ~ o

and ZSXk ~ o = <ISXk = <+ infinity

zSXk = 1 - (e EXP - ((ISXk - LSXk ~ u) / Ck))

for ZSXk ~ o = + infinity
and LSXk ~ u = <ISXk = <+ infinity

zSXk = ISXk / ZSXk ~ o for 0 <ZSXk ~ o <+ infinite

and ZSXk ~ o = <ISXk = <+ infinity

For ordinal values, the actual values ISXk can be the Degree of target achievement zSXk can be assigned personally. With a few, discrete variants, e.g. B. Pollutant classes in cars, this happens immediately, with many discrete or with continuous values this can be done indirectly through bases, e.g. B. by specifying the best values with target achievement zSXk 0.0, 0.1, 0.2 to 0.9 and 1.0 and through, if necessary mechanical, classification of the Actual values in the corresponding value ranges.

With nominal values, all hits have the Degree of target achievement 1.

2. Calculation of the number of values WBVi depending on Assessing B and its preferences.

Several evaluators B with the same search target ZSXk and search limit LSXk can evaluate the same actual value ISXk differently. The regulation VBVi records these differences for the evaluation of the degree of target achievement zSXk with the value number WBVi. In general, the formula applies

WBVi = VBVi (zSXk, evaluator (B)) with 0 = <WBVi = <1
and WBVi = 0 for zSXk = 0
and WBVi = 1 for zSXk = 1,

where the value number WBVi is standardized to values between 0 and 1 and always set to 0 or 1 for the degree of target achievement 0 or 1. The pre-assignment is suitable for calculating the intermediate values

VBVi = zSXk EXP ESXk with ESXk = 1.

This pre-assignment can be changed by the assessor B. by choosing another exponent or another function. Preferably a rating B z. B. balanced objects or amounts of data without pronounced weakness, then exponents are suitable ESXk less than 1, while objects are preferred with pronounced strengths exponent ESXk greater than 1 are suitable.

This evaluation, which takes place during the search, can be illustrated using performance data:
With variant A, the place of performance is a nominal quantity with a value of 1 for all hits. In the environment search for variant B, the distance is a scalar variable with 0 as the search target ZSXk and a finite upper search limit LSXk ~ o. The pre-assignment of the regulation VBVi can remain unchanged.

The performance period is a scalar variable with an upper, finite search limit. The advantage due to a quick The provision of services depends on the individual case. On earlier appointment in some cases results in direct and in other cases indirect benefits because with the buffer time the probability of meeting deadlines  increases. Therefore, the search target ZSXk are suitable for Value 0 and for the exponent ESXk the value 0.5.

The performance price is a scalar variable with an upper, finite search limit and search target 0. In some cases, especially for products with validity, the Rating B not only upwards but also also limit down and one between the Select search limit lying search target. The default the assessment rule can be used in all cases become.

The mediator V could be used to evaluate a Data volume per evaluation criterion BMi only one evaluator allow. This solution is not suitable for permanently existing data volumes. It is unsuitable for ratings by seekers S since the ratings turn out very different and the value numbers of the first seekers will rarely be representative.

It is therefore appropriate, depending on the amount of data and evaluation criterion BMi over time several reviews BMij, which are referred to as individual assessment and the associated value numbers WBMij, quantity numbers MBMij and times tBMij to record and at least save temporarily.

If a data set for an evaluation criterion BMi BMij receives individual ratings over time, then with their value numbers WBMij, quantity numbers MBMij and times tBMij the value number WBMi, quantity number MBMi and time tBMi of the current evaluation of the BMi criterion calculated and saved.

Different formulas can be used for the calculation. Are particularly suitable

• - for individual evaluators B, e.g. B. the provider A or the intermediary V, the formula
  WBMi = WBMim (A)
  Here the value number WBMim of the last individual evaluation BMim overwrites the previous value number WBMi. The quantity number MBMi is from the mediator V to values greater than 0, z. B. 1 set.
• - for several evaluators B belonging to the same group, in particular seekers S,
  the formulas without taking into account the valuation times
  WBMi = sum j (WBMij × MBMij) / MBMi (B1)
  MBMi = sum j (MBMij) with j = 1 to m
  The value number WBMi is a mean value of the value numbers WBMij weighted with the quantity numbers MBMij. The quantity numbers MBMij are constant over time.
  If the intermediary V sets the quantity numbers MBMij to 1, then he simplifies the formulas B1)
  WBMi = sum j (WBMij) / MBMi (B2)
  MBMi = m with j = 1 to m
• - for several evaluators B belonging to the same group, in particular searchers S. taking the evaluation times into account, the formulas
  WRMi (tBMi) = sum j (WBMij × MBMij (tBMij = 0) × tf (tBMij)) / MBMi (tBMi) MBMi (tBMi) = sum j (MBMij (tBMij = 0) × tf (tBMij)) (C1)
  with j = 1 to m
  and tf (tBMij = 0) = 1
  With these formulas, the quantity numbers MBMij are changed with the age tBMij of the individual evaluation BMij. For this purpose, the quantity MBMij (tBMij = 0) at the time of the individual evaluation is multiplied by the time function tf (tBMij). The value number WBMij as a qualitative component of the evaluation remains constant over time.
  The intermediary V expediently sets the original quantity numbers MBMij (tBMij) to 1 and uses a time function tf with values that are constant or decrease over time. That gives the formula
  WBMi (tBMi) = sum j (WBMij × tf (tBMij)) / MBMi (tBMi) MBMi (tBMi) = sum j (tf (tBMij)) (C2)
  with j = 1 to m
  and t2 <t1 = <tf (t2) <= tf (t1)
  and tf (tBMij = 0) = 1
  The intermediary could permanently save all previous m - 1 individual evaluations and recalculate the evaluation for each criterion after a further individual evaluation BMim. He can simplify the calculation if he always condenses the individual evaluations BMij into the evaluation for each criterion BMi and links this evaluation with the further individual evaluation BMim. Eg the formulas B2) then result
  WBMi = (WBMi (a - 1) × MBMi (a - 1) + WBMim × 1) / MBMi MBMi = MBMi (a - 1) + 1 (B3)
  Calculations with time-dependent quantity numbers can also be simplified in this way if the mediator uses a suitable time function tf.

When linking several individual assessments BMij, that over time for a set of data and a Assessment criterion made BMi, their age tBMij disregarded and their value number WBMij and Quantity MBMij remain unchanged. However, it is often appropriate to take age into account, since current individual assessments of the current condition and correspond to the current view, while earlier ones Individual reviews are often outdated, e.g. B. because interim improvements to the property or the amount of data or due to increased claims of  Rating B. The previous value numbers are off current view, in the first case too low and in second case too high. The reduction in quantity In contrast, numbers with their ages lead in both cases to approximate the current view.

Therefore, as described above, the original Quantity number with a time function tf, its value is initially 1 and can then decrease, multiplied.

The exponential function is well suited as a time function

tf (tBMij) = e EXP - (tBMij / D0)

with 0 <D0 = <+ infinite,
because then the evaluation for each criterion can be calculated from previous individual evaluations and linked to further individual evaluations.

With small time constants D0, the quantity numbers decrease fast, with large time constants they decrease slowly and with time constants + infinite, the quantity Numbers constant.

With the times tBMij (a) of a current individual evaluation BMij (a) and tBMij (a - 1) of the last previous individual evaluation BMij (a - 1), the period DBMi between these two evaluations applies

DBMi = tBMij (a) - tBMij (a - 1) for all j = 1 to m.

With this period of time and the previous value number WBMi (a - 1) and quantity number MBMi (a - 1), a further individual evaluation of WBMim and MBMim results in the current value number WBMi (a) and quantity number MBMi (a) the formulas

WBMi (a) = (WBMi (a - 1) × MBMi (a - 1) × (e EXP - (DBMi / D0)) + WBMim × MBMim) / MBMi (a)

MBMi (a) = MBMi (a - 1) × (e EXP - (DBMi / D0)) + MBMim.

If the value numbers of current individual evaluations clearly differ from previous individual assessments and mostly better or mostly worse then this is an indication of improvements in the Objective, rapidly increasing demands of the evaluators B or fashionable change.

Therefore, depending on the amount of data and evaluation criteria constantly checked whether changes in the number of values compared previous value numbers are significant, taking note the quantity numbers and the times of the evaluations. If they are significant, then previous quantity Numbers additionally reduced to a faster adjustment the number of WBMi values based on current individual assessments To reach WBMij.

The current value number WBMi and quantity number MBMi des Criterion BMi could be added after each additional BMij rating can be calculated. However, the number of the calculations reduced and the above significance test can be made easier by using Time intervals DI and summarizing the inside individual evaluations carried out at a time interval.

It is advisable to leave the previous value number WBMi (a - 1) and quantity number MBMi (a - 1) unchanged until the end of the first following time interval with at least one individual evaluation, and then the quantity number per elapsed time interval with the time factor tF, which according to the formula

tF = e EXP - (DI / D0)

is calculated to decrease.

The time span between the current time interval a and the time interval (a - 1) with the last individual evaluation BMim-1 is a multiple Ci of the time interval DI. The following applies to Ci

Ci = DBMi / DI.

It is also expedient to define the individual evaluations made in a time interval as being of the same age and to set their evaluation times uniformly at the end of the respective time interval. Then the equation applies to the mean value number WBMi, a of the n individual evaluations BMij in the current interval a

WBMi, a = sum (WBMij) / n with j = m to m + n - 1.

If the time interval DI is selected appropriately, this contains current interval a is usually a sufficient large number n of the individual evaluations BMij, so that the Significance of changes in the number of values with the values for WBMi (a - 1), WBMi, a, n, Ci and DI can be checked.

The following then applies to the current value number WBMi and quantity number MBMi

WBMi = (WBMi (a - 1) × MBMi (a - 1) × (tF EXP Ci) + WBMi, a × n) / MBMi

MBMi = MBMi (a - 1) × (tF EXP Ci) + n.

It should be noted that the search results Value numbers immediately added to the selected amount of data be arranged and until the hit list is closed remain unchanged.

The judge B receives after creating the hit list a connection to the first hit or to the am Amount of data at the beginning of the hit list. The further contact is made by entering the assessor B determines and takes place in stages that are called contact stages be designated. The contact levels enable Ent divorce-oriented assessments E of the inputs of the Rating B.

The mediator can e.g. B. generally use the following contact levels

• - Representation of a 1st part of hit X
• - Display further data of hit X  
• - Establish connection with hit X.
• - 1. Send data to hit X
• - 2. Send data to hit X after it answer

and start with the third contact level for immediate dialing.

The mediator lets each hit increase the Contact level, reducing the contact level and that Repeating these processes several times for hit X or for other hits too. He can do his analysis for hit X except the contact levels z. B. also the Determine times per contact level or the processing Consider other hits in the hit list.

The mediator V can for the seeker S as Rating B z. B. define the following 8 criteria SEi. Does he use these criteria for all amounts of data, then he can compare any amount of data.

The data determined for evaluators B are evaluated by decision-oriented evaluation E according to

WBEij = VBEi (for BEij determined data),

In addition to these implicit individual evaluations, too explicit individual ratings BPij by the evaluator B recorded based on personnel rating P.

Provider A strives to exchange its services Seeker S. His chance of success is determined by the frequency with which its data set is selected and by the sorting location within the selection. He can increase his chance of success by using the Da quantity according to the selection and sorting algo rithmen designed or rated very well. He can too have more data in the database.

As long as there is no disadvantage to this advantage, provider A as assessor B is not neutral and not suitable. However, if the higher chance of success at least partially offset by a disadvantage then an evaluation by the provider A useful and their review by the intermediary V for abuse, e.g. B. Word spamming on Internet Search engines or AAA at the beginning of the company name printed directories, not required.

The mediator V therefore forms evaluation criteria AVk for the evaluation according to the regulation by the provider A and enables provider A to Factor WAVk, which is the value number WAVk and at the same time the Amount of a contingent disadvantage of the provider A determined to fix.

The two general formulas apply to the evaluation by provider A by means of a limited disadvantage

WAVk = VAVk (wAVk, reference value)

Disadvantage = f (wAVk, value for reference quantity, occurrence of the condition)

If the mediator V the conditions of disadvantage match the reference size, then the advantage better chances of success, e.g. B. by better sorting places, at a disadvantage, e.g. B. a higher payment amount coupled.

The reference variable can be fixed, one fixed in periods or be a variable size, e.g. B. Recording a Da amount in the database, save the amount of data for a year or output of the amount of data within one Mediation to a seeker. Before as a reference sizes are added, their values for all data quantities immediately available or for other reasons be determined, e.g. B. Value numbers WSEij for seekers S and decision-oriented assessment E.

Variable reference values, e.g. B. WSEij, motivate the Provider A, the data of his data volume is correct to enter and quickly to his current situation adapt, e.g. B. by lowering the value factor wAVk 0 at high capacity utilization. The provider A avoids disadvantages and improves at the same time the quality of the hit list because then others Moving data volumes forward.

Provider A chooses a high value factor if the experienced or expected benefits of the performance exchange is great with seekers S. Because the power output exchange only with the consent of the seeker S, ent this rating speaks to the interests of seekers S.

The mediator V can differentiate the value Allow WAVk factors, e.g. B. by performance zone OAL, Order entry time tAA, performance time tAL, performance number of ASL or search data. When differentiating Provider A can search data for potential new customers make a high payment bid, since seekers S. the him already know and use the convenience of instant dialing, will enter his name.  

The mediator V can e.g. B. ask for a bid, request a fixed fee or, in exceptional cases, access providers without a payment bid and the value factors wAVk for an evaluation criterion AVK

• - let providers A choose freely or
• - one for each object class for all data quantities set the same value or
• - 0 for each object class or one for all data set the same value and the providers only leave the choice of one of these two values.

A payment amount works well as a disadvantage as it is a scaled size and from all providers is actually perceived as a disadvantage. Another The advantage is that the associated income on several Evaluating B, also to seekers S, and be distributed can compensate for their effort.

Under these conditions, the two general formulas can be replaced by

WAVk = wAVk × VAVk (reference value)

RAVk = wAVk × f (value for reference variable, condition occurs)

with RAVk as the invoice amount, for the payment of which Provider based on the value factor he has defined and is subject to conditions.

For standardized value numbers, e.g. B. WAVk with value range 0 to 1, it makes sense to normalize the value factors and the reference values, or to use standardized reference values. Then the equations apply

WAVk = wAVk / wAVk ~ max × VAVk (standardized reference value)

RAVk = wAVk × (value for standardized reference value) × f (condition occurs)

with wAVK ~ max as the maximum value factor of the criterion AVk, preferably per object class, and that on 0 to 1,  or normalized to 0 and 1 for bipolar quantities Reference.

The WBMij value numbers are particularly suitable as standardized reference values. The condition of the provider disadvantage is then the success of an individual assessment BMij. So that applies

WAVk = wAVk / wAVk ~ max

RAVk = wAVk × sum j (WBMij)

RAV = sum k (RAVk)

The amount RAV that provider A has for a set of data pays, results from all evaluations of the criteria AVk. He can the provider z. B. Invoiced weekly be placed and debited.

Suitable conditions of disadvantage for provider A are disadvantages of other evaluators B, z. B. effort of Time and / or money through the intermediary V, seeker S, Expert G or less chance of success for others Amounts of data due to poorer sorting locations.

At intermediary V, the input and output of data increases the required storage and computing capacity and thus the fixed costs.

For seekers S, the main disadvantage is the amount of time z. B. for checking amounts of data due to miss customary design as part of the hit list issued, checked by seeker S and not specified clicks or clicked by entering a confirmation and were discarded after further data was output.

Reviewers G need for the preparation of reviews Time. This effort can e.g. B. from the intermediary V to the Provider A can be apportioned.

Worse sorting places of other data sets are in usually without any effort. The associated disadvantages  but are also a reason, part of the benefits skim off provider A.

A new amount of data is added to the database, by provider A creating them. If the provider its data volume is not evaluated, then the new one would be Amount of data without evaluation and the total value number W not defined. This also applies to those outside Datasets of data stored, their location is known and recorded internally for the first time.

Therefore, the intermediary V rates new amounts of data with this mentioned starting values. He avoids the effort Personnel ratings and machine rated. There however other ratings, e.g. B. the evaluation by Seeking S. result in more reliable values, uses the Mediators the starting values only until reliable Reviews are available. He strives for meaning the start values automatically decrease and assign start values preferably for criteria where short term Reviews are expected.

So good value numbers, e.g. B. 0.9, for WSEi or WSEij together with low quantity numbers, e.g. B. 10, for MSEi for decision-oriented assessment E by Searching S good sorting places and thus a high one Probability of evaluating the data sets become. After these assessments, data sets that are rated well by seekers, still good ones Sorting locations while badly rated data sets due to the low quantity number quickly on rear Sorting places are displaced.

The total value number W determines the order of the Hits in the hit list. It is calculated from the Figures WBMi of the evaluation criteria BMi calculated. The calculation can be done in partial steps taking into account the groups BM and B of the Evaluation criteria BMi.  

It is usually sensible that the total value W is not as arithmetic mean, but as a weighted mean to calculate the value numbers WBMi and thereby the Weight figures GBMi of the criteria BMi must be observed. There the data sets contain criteria without evaluation the quantity MBMi is also taken into account. at The criteria with quantity number 0 becomes the weight number GBMi set to 0.

It is advisable to set the value numbers to the value range 0 to normalize to 1, so that z. B. a value number of 0.9 always corresponds to a very good rating. meaningful is also to standardize and weight the GBMi due to the weighted normalized to a sum of 1 To replace numbers gBMi.

The weighted average W is calculated with the value numbers WBMi and the weight numbers GBMi all criteria BMi taking into account their Quantity MBMi.

In the case of an additive link that allows value numbers less than 0, the total value number W is calculated with

W ~ additive = total BMi (WBMi × gBMi)

Multiplicative links are also possible, which do not allow value numbers less than 0, according to the equation

W ~ multiplik. = Product BMi (WBMi EXP gBMi)

and mixed forms of additive and multiplicative Shortcut.

The standardized weight numbers gBMi are calculated using the equations

GBMi = 0 for MBMi = 0
GBMi = GBMi / G for MBMi <0
G = total BMi (GBMi) for MBMi <0.

The prerequisite for this calculation is that every hit contains at least one criterion BMi, the quantity MBMi number and its GBMi weight number greater than 0 are. This can be ensured by starting values.

It should be noted that even with data volumes Structure different evaluation criteria active could be. Therefore, the standardized weight numbers gBMi be different from hit to hit.

The assignment of evaluation criteria to groups of Criteria and target values of the non-standardized Weight numbers on these groups improved that Comparability and is also suitable for unstructured tured search, its hits to different Object classes with not comparable criteria can belong.

The assignment according to is suitable for group formation Evaluation method M and evaluation B. Main groups B per evaluator B with subgroups BM each Evaluation method M formed. More main or Sub-groups can be formed, e.g. B. for Evaluation criteria with a particularly high number of weights, z. B. Place of performance and price.

With the same target values GBM ~ target of the weight numbers GBM of the groups BM that apply to all hits, the standardized weight numbers gBMi per hit and criterion BMi are calculated using the formulas

MBM = Sum i (MBMi)

MB = sum M (MBM)

G1 (BM) = GBM ~ Soll

G2 (BM) = GBM ~ target for MBM <0

G2 (BM) = 0 for MBM = 0

G1 (B) = sum M (G1 (BM))

G2 (B) = sum M (G1 (BM)) for MBM <0

G2 (B) = 0 for MB = 0

G2 = sum B (G1 (B)) for MB <0

G3 (BM) = Sum i (GBMi) for MBMi <0

G3 (BM) = 0 for MBM = 0

gB = G1 (B) / G2 for MB <0

gB = 0 for MB = 0

gBM = G1 (BM) / G2 (B) × gB for MBM <0

gBM = 0 for MBM = 0

gBMi = GBMi / G3 (BM) × gBM for MBMi <0

gBMi = 0 for MBMi = 0

The communication network can e.g. B. as a radio network, be wired wirelessly or as a cable network. Combinations of both are also possible Network types for at least one, the data or the Part of the transmission link used.

The functions of the various means of communication partly overlap. How computers enable especially multimedia PCs, also phone calls and Telephones, e.g. B. WAP mobile phones, also the transfer of Data. Even modern TV sets cannot do data only receive, but also send.

The mediator V preferably only starts A means of communication and integrates Internet and Phone. The phone numbers and web addresses shown are then not comment fields, but links.  

The change in the means of communication is avoided. The connection between evaluator B and intermediary V remains closed until the evaluator B uses it completed. Since the evaluating numbers, e.g. B. Telephone numbers, no longer to be entered personally comfort is improved. simultaneously the intermediary V receives further possibilities that Evaluate the decisions of the evaluators B and the Improve the quality of the hit list.

The above Some means of communication have visual input means, e.g. B. Barcode scanner. Audi Elle Inputs enable microphones, tactile inputs e.g. B. Keyboard, pointing device or more touch sensitive Screen. Outputs are made visually using an image screen or display and audibly via loudspeaker. Some means of communication allow tactile output, z. B. Vibration alarm.

After this description of how the Switching system, the invention is based on two special embodiments in more detail are explained.

Take the taxi center, for example

The agent V also arranges taxi rides Object class "Taxi" created and input masks for Object master data, performance data and search data created.

A taxi operator with taxi vehicle Tx wants to Provider A contacts to passengers as seekers S get mediated. He selects as using a PC Means of communication and the Internet as communication network the website of the Mediator V. The mediator puts one on him Mask for provider master data and other input masks  ready. The structure of the masks for the providers master data, the object and performance data are in the Tables 1 to 3 are shown.

The intermediary checks the information on companies and Vehicle Tx, assigns identification numbers and authorizations and unlocks company and vehicle for mediation.

A passenger selects the using a radio telephone and radio network Telephone number of the intermediary V for seekers S. Er receives a start mask for entering the object class and then an input mask tailored to the object class, which is explained in more detail by Table 4.

The passenger as the seeker gives S as the object class "Taxi" and usually confirms the pre-receipt Ungen.

The mediator V checks the operationally reported Taxis, selects suitable taxis and sorts them their total value number. By means of direct dialing, the Passenger connected to the radio of the 1st taxi.

Normally, the passenger orders the 1st taxi. Then the passenger closes the connection with the 1st taxi, arranging a taxi ride and connecting to Mediator. The driver reports that he is not is ready for operation.

In special cases, e.g. B. if the driver is rude, The passenger does not order the 1st taxi and closes just connecting with this taxi. The passenger will then immediately with the second and possibly further taxis connected until he closes the operator.

The intermediary's internal processes and algorithms for selection, evaluation and sorting are explained in more detail in Table 5:
When selecting suitable taxis, intermediary V compares the object class, then the driving area or city and the operational readiness of the taxi. For the distance between the passenger and the vehicle, he uses the value + infinity for the associated search limit LSX1, which does not exclude a taxi, and for the degree of achievement zSX1 an exponential function with the constant C1 of 2.885, so that 2 km distance gives the value 0.50.

He rates and sorts the selected taxis considering the distance ISX1, the payment bid wAV1, which refers to the exchange SE5, and the rating SE8 by passengers as seekers S.

During the mediation, the mediator V analyzes the Processing of the taxis by the passenger. The procedure this is illustrated by Table 6.

The intermediary V then calculates the payment obligations of provider A arising from the mediation:
Evaluation criterion: AV1 bid;
Payment obligation: RAV1j = wAV1 × WSE5j × MSE5j = DM 1.50 × 1.00 × 1 = DM 1.50.

The sum RAV1j results in the payment per period.

The passenger evaluates the driver during the call. Most of the time he only needs a taxi and calls other taxis only if he takes the 1st taxi or whose driver is not satisfied. This single rating The intermediary V links the SE8j of the taxi with the Rating SE8 by previous passengers to the new one Review SE8 ~ new. Like the calculation after every call Table 7a illustrates.

The newly approved taxi Tx has initially no rating SE8 by passengers. Since the Taxi operator did not commit to payment,  the value number WAV1 = 0.00. The mediator V would like have the taxi rated by passengers and the Taxi entrepreneur A motivate yourself to pay commit. Both require the mediation of Passengers S to the taxi Tx.

The mediator therefore arranges good amounts of new data Starting values too, for taxis, for example, for the criteria rium SE8 in the form shown in Table 7b.

The taxi Tx receives on the 1st day based on these starting values multiple calls from passengers. The driver is not used to speak to passengers and unfriendly Lich. A large part of the potential passengers ends the call and leaves with the next one Connect taxi. The first day is:

For a rating with:

• - Distance: 1.00 km = ISX1
• - Payment bid: 0.00 DM = wAV1 and
• - Rating: 0.90 = starting value

the following value numbers result:

• - WSX1 = e EXP - (1.00 / 2.885) = 0.707,
• - WAV1 = 0.00 / 1.50 = 0.00 and
• - WSE8 = 0.90,

which to the total value number
W = 0.707 x 0.48 + 0.00 x 0.32 + 0.90 x 0.32 = 0.627
be compressed.

In the evening the taxi received 11 calls but only 3 orders, so the rating SE8 gets worse:
"Good":
WSE8j = 1.00 for j = 2, 7 and 9
MSE8j = 1
"Bad":
WSE8j = 0.00 for j = 1, 3 to 6, 8, 10 and 11
MSE8j = 1

For the new evaluation of criterion SE8, this results in:
WSE8, a = (0 + 1 + 0 + 0 + 0 + 0 + 1 + 0 + 1 + 0 + 0) / 11,
WSE8, a = 0.273 WSE8 = 0.90,
MSE8, a = 11 MSE8 = 10 so,
WSE8 = (0.90 x 10 + 0.273 x 11) / 21 = 0.572
MSE8 = 10 + 11 = 21.

The driver is dissatisfied with this mediation. The In contrast, intermediary V analyzes the processing of the Hit, automatically creates an evaluation and sends by taxi operator A on the evening of the first day E-mail. He points out that 11 inquiries were conveyed that 8 passengers the order processing refused and that when entering into the obligation to pay 17 requests would have been brokered.

The taxi operator then makes the payment commitment and drives himself on the second day. He is polite and receives almost all orders. For the second day:
For a rating with:

• - Distance: 1.00 km = ISX1
• - Payment bid: 1.50 DM = wAV1 and
• - Rating: 0.572 = from the previous day

the following value numbers result:

• - WSX1 = e EXP - (1.00 / 2.885) = 0.707,
• - WAV1 = 1.50 / 1.50 = 1.00 and
• - WSE8 = 0.572,

which to the total value number
W = 0.707 x 0.48 + 1.00 x 0.32 + 0.572 x 0.20 = 0.774
be compressed.

The following situation would result without a payment bid:

• - Distance: 1.00 km = ISX1
• - Payment bid: 0.00 DM = wAV1 and
• - Rating: 0.572 = from the previous day

and thus the value numbers:

• - WSX1 = e EXP - (1.00 / 2.885) = 0.707,
• - WAV1 = 0.00 / 1.50 = 0.00 and
• - WSE8 = 0.572,

which to the total value number
W = 0.707 x 0.48 + 0.00 x 0.32 + 0.572 x 0.20 = 0.454
be condensed so that the provider would hardly have received any orders without a payment offer because of the poor rating SE8.

At the end of the second day, the provider took 13 calls and received 12 orders:
"Good":
WSE8j = 1.00 for j = 12 to 13, 15 to 24
MSE8j = 1
"Bad":
WSE8j = 0.00 for j = 14
MSE8j = 1

For the new evaluation of criterion SE8, this results in:
WSE8, a = 0.923 WSE8 = 0.572,
MSE8, a = 13 MSE8 = 21 so,
WSE8 = (0.572 × 21 + 0.923 × 13) / 34 = 0.706
MSE8 = 21 + 13 = 34.

The value number WSE8, a of the second day at 0.923 is significantly better than the previous value number at 0.572. A significance test is carried out automatically and then the previous value of the quantity number MSE8 is reduced, e.g. B. on the quantity of the starting value:
WSE8, a = 0.923 WSE8 = 0.572,
MSE8, a = 13 MSE8 = 10 and thus,
WSE8 = (0.572 x 10 + 0.923 x 13) / 23 = 0.770
MSE8 = 10 + 13 = 23.

This measure results in a faster adjustment of the Value number WSE8 to the current state.

Example 2 Two companies with place of performance, performance zone and performance time

The example shows how changes to the object and Performance data, the sort order and the placement determine. Two providers are compared: One first provider, provider ~ 1 of his inputs leaves unchanged, and a second provider, provider ~ 2, who changes his entries over time.

Provider ~ 1 is a large copy company with 30 Mit workers in the city center of Leipzig with Postleit number 04103. He has 10 courier vehicles of his own on Mondays open and reports from Friday to 8:00 a.m. to 8:00 p.m. on December 1st, 2000 at agent V.

Important data resulting from the registration of the Providers ~ 1 result are shown in Table 8.

Provider ~ 2 is a small copy company with 3 Employees in the south of Leipzig with the zip code 04277. He has a company vehicle that is also for couriers trips is used Monday to Friday from 8:00 open until 6 p.m. and reports on January 2nd, 2001 at the agent.

Important data resulting from the registration of the Providers ~ 2 result are shown in Table 9.  

01.02.2001: Provider ~ 2 extends the opening time to 10 p.m. (*) and also opens on Saturdays. The delivery service does not change provider ~ 2 (**). However, it increases the Value factor of 3.00 (***) (see Table 10).

02/27/2001: Vendor ~ 2 is busy, as new customers place all orders in its operations. He understood the selection and sorting algorithm and improved the value factor and the offer (see Table 11):
At 07:30: He lowers his value factor (*), remains just above the bid of provider ~ 1 at 2.10 and chooses the value 1.00 for Saturday. He uses taxis (**) for urgent orders, since a taxi is always close to the customer and the trip from the company to the customer to collect the templates is not necessary. The transport time is reduced from 1.8 to 0.8 h due to the reduced number of journeys to 2 and the 0.4 h shorter journey time for taxi drivers. The delivery time DAE (***) shortens providers ~ 2 by interrupting other orders when rush orders arrive.

At 4:30 p.m .: Tomorrow is the closing time for an architect Competition. Architects want "a few plots" quickly. Provider ~ 2 is fully loaded 02394 00070 552 001000280000000200012000285910228300040 0002010107792 00004 02275s and already has 4 Have to reject requests. Now he is deleting plot Service temporary (****).

The mediator V uses photo, Color and large copy as well as plot service the same Selection and sorting algorithms. He demands from the Providers the specification of a value factor wAV1, which is refers to switch SE5 and gives them the Possibility to choose wAV1 freely and according to object class, Differentiate between performance zone and performance time.

When making a selection, the mediator checks everyone Amounts of data of the searched object class, whether the  Place of performance OSL of the seeker within a Performance zone OAL of the provider is and whether that of Seekers desired service time tSL from the provider can be realized.

He sorts the hits selected in this way Total value number W, which in this case consists of only one Evaluation, namely AV1, is determined and therefore with WAV1 matches.

If no amount of data meets the conditions OSL and tSL, then the mediator looks for approximate solutions, e.g. B. with environment search. There is no need for these mediations the provider's obligation to pay.

The development of the search data, the resulting Sort order and mediation for the second execution Example should be for inquiries in the period 01.12.2000 until 28.02.2001 for the city of Leipzig. The corresponding procedure is in Tables 12a and 12b shown. The columns under 'Mediation' indicate the Number of the selected providers, 1 or 2, and their Value factor wAVk in DM.

The invention is not limited to the Darge here presented embodiments. Rather, it is possible by combining and modifying the means mentioned and features to realize other design variants ren without departing from the scope of the invention.  

attachment

List of abbreviations of terms and variables

Claims (32)

1. System for establishing contacts between providers (A) and seekers (S) via intermediaries (V) using communication networks, characterized in that
Amounts of data stored in at least one database
are evaluated in such a way that value-number and / or quantity-number and / or weight-number are determined for at least one evaluation criterion contained in the data set and these are compressed to a total value-number or if a data set only contains one value-number has, this value number is adopted as a total value number,
selected according to the specifications of an evaluator (B), recorded in a hit list sorted by the total value, and the contact between evaluator (B) and at least one selected data set or the provider (A) of a data set
is automatically created by immediate selection or
realized after submitting at least a part of the hit list to the evaluator (B) according to individual decision
becomes. 2. System for establishing contacts between providers (A) and seekers (S) via intermediaries (V) using communication networks, characterized in that
Amounts of data stored in at least one database
are evaluated in such a way that value-number and / or quantity-number and / or weight-number are determined for at least one evaluation criterion contained in the data set and these are compressed to a total value-number or if a data set only contains one value-number has, this value number is adopted as a total value number,
selected according to the specifications of an evaluator (B), recorded in a hit list sorted by the total value and the contact between evaluator (B) and at least one selected data set or the provider (A) of a data set
automatic immediate dialing or
Transmission of at least some of the selected data to the evaluator (B)
will be produced. 3. System according to one of claims 1 or 2, characterized in that
object data describing an object and / or
provider data describing the provider (A) of an object and / or
performance data describing the performance of at least one evaluator (B) and / or
at least one evaluation of the object data and / or provider data and / or performance data
Part of a data set and / or can be used as search data for the selection of the data sets. 4. System according to claim 3, characterized in that
through the performance data at least
Place of performance and / or
Performance time and / or
power Price
is described. 5. System according to claim 3, characterized in that
through the search data at least
Search limit and / or
Search condition and / or
search target
is described. 6. System according to one of claims 1 to 5, characterized in that
The selection of the data sets is based on
at least one actual value of the data volume and / or
the search data. 7. System according to any one of claims 1 to 6, characterized in that
the amounts of data and / or further data and / or algorithms
from the intermediary (V) and / or
from third
are included in at least one of the databases. 8. System according to any one of claims 1 to 7, characterized in that
the input of data by the evaluator (B) either
by at least one action by the assessor (B) or
by refraining from an action or
by failing to take any action within a period of time
he follows. 9. System according to any one of claims 1 to 8, characterized in that
the data sets can be assigned to an object class and each data set assigned to an object class
a structure specified by this object class and / or
has at least one same evaluation criterion. 10. System according to any one of claims 1 to 9, characterized in that
at least one input mask with at least one data field as
freely writable input field and / or writable input field for a criterion and / or
editable graphical input area and / or pull-down menu and / or
Slider for changing parameters and / or
switch
provided. 11. System according to claim 10, characterized in that each object class has at least one special one Input mask is assigned. 12. System according to one of claims 10 or 11, characterized in that Data pre-assigned to at least one input mask are. 13. System according to claim 12, characterized in that changes made by evaluators (B) Preassignments are analyzed and based on the results of these analyzes selection and / or evaluation Algorithms and / or pre-assignments changed become. 14. System according to any one of claims 1 to 13, characterized in that
the evaluator (B)
a provider (A) or
an expert (G) or
a seeker (S) or
an intermediary (V)
is. 15. System according to any one of claims 1 to 14, characterized in that
as valuation method (M)
a personal assessment (P) or
an assessment according to regulation (V) or
a decision-oriented assessment (E) is used. 16. System according to any one of claims 1 to 15, characterized in that
When searching for data fields with a search target, depending on the selected data set, the actual values of the data set and search data
a degree of goal achievement and / or
each a value number
be calculated. 17. System according to one of claims 1 to 16, characterized in that the evaluation of at least one evaluation criterion by an assessor (B) in the form an individual assessment takes place. 18. System according to claim 17, characterized in that for at least one evaluation criterion Amount of data at least one individual assessment and / or the previous evaluation of the Evaluation criteria are linked and thereby a new evaluation of the Generate evaluation criteria.   19. System according to any one of claims 1 to 18, characterized in that
the quantity of a rating is changed depending on
their age and / or
Value numbers and / or quantity numbers and / or the age of ratings for the same rating criterion. 20. System according to any one of claims 17 to 19, characterized in that
the individual evaluations of an evaluation criterion
assigned to certain time intervals and / or defined as the same age if they belong to a time interval
become. 21. System according to one of claims 1 or 3 to 19, characterized in that
the processing and / or evaluation carried out by the evaluator (B)
the hit list and / or
the amount of data contained in the hit list and / or their providers
is analyzed and, based on the results of this analysis, at least one quantity of data is reevaluated for at least one evaluation criterion. 22. System according to one of claims 2 to 20, characterized in that the processing carried out by the evaluator (B) and / or evaluation of contacts analyzed and based on the results of this analysis at least a data set for at least one evaluation criterion is reassessed. 23. System according to one of claims 1 to 22, characterized in that a provider (A) at least one rating of one performs the amount of data offered by him by the amount of a contingent disadvantage of the provider (A) indicates. 24. System according to claim 23, characterized in that the disadvantage of being charged to the provider amount is. 25. System according to any one of claims 23 or 24, characterized in that
Condition of disadvantage
enabling contacting or
contacting or
refusing to contact or
the conclusion of a contract
an evaluator (B) with the amount of data evaluated in this way and / or their provider (A). 26. System according to any one of claims 23 or 24, characterized in that
Condition of the disadvantage was that for the amount of data evaluated in this way and / or its provider (A)
Enter data with or
Transmission of data to
is an evaluator (B). 27. System according to one of claims 1 to 26, characterized in that for each new one added to the database Data volume at least one evaluation with start values for the value and / or the quantity number is provided. 28. System according to one of claims 1 to 27, characterized in that the total value number of a data set weighted average of the value numbers in the Data set contained evaluation criteria. 29. System according to one of claims 1 to 28, characterized in that Evaluation criteria grouped together and each group of evaluation criteria a weight number is assigned and this Groups one or more by the same Rating (B) and / or after the same Evaluation method (M) Evaluation criteria included.   30. System according to one of claims 1 to 29, characterized in that
the communication networks
Radio networks and / or
cable networks
are. 31. System according to any one of claims 1 to 30, characterized in that
the input and / or output of data by means of communication by the evaluator (B)
visual and / or
auditory and / or
tactile
Means. 32. System according to claim 31, characterized in that
the means of communication
Computer and / or
Phones and / or
televisions
are.