WO2002054709A2

WO2002054709A2 - Information search, information matching and information provision based on dynamic and address systems in networks in the form of content dependent dynamic dns (domain name service)

Info

Publication number: WO2002054709A2
Application number: PCT/DE2002/000001
Authority: WO
Inventors: Dirk Liesch; Janko Nebel
Original assignee: Dirk Liesch; Janko Nebel
Priority date: 2001-01-06
Filing date: 2002-01-03
Publication date: 2002-07-11
Also published as: EP1410604A2; DE10100229A1; WO2002054709A3

Abstract

1. A method and technical system for dynamic information routing in distributed networks. 2.1 Target-oriented dynamic automatic routing of information requests to the information sources pointing in distributed networks by transparent use of the DNS protocol. The above permits significantly more effective and direct information access than with conventional search engine technology. 2.2 A dynamic address assignment of information sources using the domain identification from the third level domain is achieved by the replacement of current DNS servers whilst retaining the DNS protocol. The modified LUAR DNS service thus delivers meaningful addresses according to the knowledge base and application example, independent of whether the above are different return addresses for the identical request, or whether the most likely desired address for an unknown request is delivered. 2.3 The method and system are particularly suitable for localised, personalised, decentralised information networks (Internet, Intranet, eCommerce), search services for application therein and planned load balancing .

Description

Title of the invention:

Information search, information matching and information provision based on dynamic routing and address systems in networks - Content-dependent dynamic DNS (Domain Name Service) - Local Universe Address Resolution (LUAR)

field of use

This invention relates to a method and a technical system for address resolution of information, and this dynamically and content-specifically for a formation request with the aim of establishing distributed information networks. The invention is used to determine addresses of information in order to enable dynamic routing of information flows and targeted access to information. For this purpose, various options for information search, matching and information provision are combined. This results in the above-mentioned title of the invention.

Any abstract information (e.g. pizza) receives an identical or similarly structured address in the application of the invention, such as a computer on today's internet. The type of information addressing according to this invention is an alternative method for currently addressing content on the Internet by means of URL. The new system can also be used in addition and together with the URL addressing. The invention also proposes some specific implementations in the field of application of distributed information clusters and fuzzy information and target terms.

The technical starting point of the system is the standard protocol used today for determining computer addresses in TCP / DP networks DNS (Domain Name Service). The basic expanded view of the new system is that the information society is looking for an address of an "information". In other words, there is a need for information and the address is sought for the appropriate know-how carrier (wise men) who can cover the need for information To this end, the information seeker expresses his wish in his own conceptual world (his names) and according to his knowledge, and then receives from this system the address (or addresses) of the "wise men" who are familiar with it. As with today's DNS with computer addresses, this can be done completely transparently for the end user. He therefore does not notice anything of this address request of his information request, but is directed directly to the "wise man" (to his address). This is done in the same way as in today's DNS, in which the connection is established with the computer whose name you have entered ,

The setting up of the first implementation variant of this method on the existing DNS protocol is an essential part of this invention, since it can be used directly with the introduction of this globally installed DNS protocol already on the market. This ensures complete transparency for the client side, which means that all users of DNS (eg all users of the Internet) are available to this system from the beginning as potential users of the system. The new system in implementation variant 1 (LUAR version 1) is thus used like a standard DNS server, only that these server addresses determine "information requests" instead of computer names. This is possible due to the fact that the saying aptly describes: "Names are just smoke and mirrors". It is not possible to determine whether the name of a DNS request is actually a computer name as provided in the protocol and can therefore be any name and thus also describe any information. The extended implementation variant 2 (LUAR version 2) requires an expansion of the DNS protocol with functionality for localization and personalization in order to be able to process these resulting dependencies transparently in the DNS for all other systems. In implementation variant 1 (with standard DNS protocol), extensions of the functionality of the "wise men" are required for localization and personalization, or the information request must be supplemented with the necessary parameters in accordance with DNS, which increases the user's knowledge or access requires an upstream application (e.g. HTML page). The invention exemplifies two technical forms as systems. Firstly, the implementation as a single system, in which the functionality is implemented within a single system, and secondly, the implementation via the basic system and a separate knowledge management system (Knowledge Base), which is an exemplary example for the realization of the invention by means of distributed knowledge management systems.

Due to the dynamic address assignment by means of knowledge management systems, the invention also enables the establishment of a targeted long-term load distribution (load balancing) without complex additional hardware in the information services and without redirect methods that are frequently used today.

The invention can be used in addition or in combination with other technical systems for information search and provision, e.g. Search engines of the Internet and HTTP request based procedures are used.

Disadvantages of the prior art

The Domain Name Service (DNS) is currently used on the Internet to determine a computer's network address (IP address) from its domain name. This is a global standardized protocol which, based on a hierarchical server structure, allows worldwide and perfoπnant the determination of an IP address for a certain computer name and vice versa the determination of the computer name to the IP address. However, today's technical DNS systems are based on a static resolution of computer names to IP addresses. For each name (eg "garfield.dlconsult.de"), a fixed IP address must be specified in the configuration in the configuration file (zone file) in the DNS server. This means that the computer name must be defined and configured beforehand. The system is currently only used to determine computer addresses for computer names and domain names and not to determine "information addresses" in the sense of the invention, l the lowest level of the DNS hierarchy (the server which provides the final address) is the static mode of operation for the scope of the invention An exclusion criterion in knowledge management in distributed networks, if one considers the computer address as a "knowledge address" and equates the computer name with a "knowledge designation". For such a purpose, a current DNS system would require an exact, clearly predetermined knowledge designation in the "zone file" in order to determine the "know-how" bearer (knowledge address), which does not allow any useful use in the area of application. The current DNS systems are therefore not suitable for solving this changed task. Your technical limit is in the case of static domain names such as "All the best for Christmas." Reached.

According to the current state of the art, the address of information is derived from its URL (Uniform Resource Locator), which is basically based on the view that hiformation resides on a computer and can be accessed there via a specific access. As a result, a URL consists of <type identifier> <computer name> <information path> <information>. From the new point of view of the invention, it does not make any sense at all to know a “computer name” in order to receive information. Why does one have to know a computer name for information? The invention assumes that this is not necessary and is too complicated for users Therefore, the invention realizes a new way of the "information address" on the basis of common designations of information using the standard for designations of domain names as syntax for designation input.

The URL always designates exactly one "information unit" (eg HTML page on the Internet) and no knowledge carrier "Wise" who knows about the area of knowledge. This means that the likelihood of finding the information you really need with a URL is not very high, unless you know from previous calls or information (additional information) that exactly the information you want can be found there. Every further attempt to find the desired information may require completely different URLs, which result from no logic. This dilemma has led to the emergence of today's Internet search engines. They try to compile the possible URLs for a content accordingly. This access via search engines, which can also be regarded as a kind of "way", represents the current search for the system of the information address according to the scheme URL. However, the user must have the address of the first one You already know "Wise" (the URL of the search engine) and go to it. Then the user has to correctly formulate his query and wait for the results and evaluate them himself. Then the user selects the hopefully correct URL from the list of results and then ends up with the desired one This procedure currently leads to the fact that the really required information in a distributed information network (eg Internet) based on URL and search engines can only be found by ever fewer people with increasing effort The state of the art in the provision and presentation of dynamic content, as well as in load distribution and performance enhancement in information services on the Internet, is based on what are known as HTTP request-based methods the total e URL including some parameters of the request are available. These methods have in common that the corresponding knowledge carrier "Wise" has already been addressed at this point in time. The query is therefore already directed to a computer in the information network and cannot be made generally to the distributed system (disadvantage of this prior art), such as this is done by means of the invention with LUAR. This is to be equated with the "wise men" according to the conceptual world of this invention. This "manner" is now able to provide the information with the help of such known HTTP request-based methods and thus represents an example of a combination of LUAR with existing methods in the overall network of a distributed information system. In concrete terms, these HTTP request-based methods include Technology of the "virtual" servers of all common HTTP servers, some methods of load balancing using HTTP requests, methods of request redirection (eg to accelerators such as the "reverse squid") and methods of personalization and localization based on the evaluation of parameters, Parts and properties of the HTTP request Furthermore, HTTP request-based systems have a "bottleneck" or "single point of failure" due to the technical system already addressed via HTTP request, which in terms of scalability and reliability even in systems from a single source Hand (an information provider) negat iv can impact. These systems are particularly unsuitable for information services of any information provider distributed worldwide using any technical systems for the "wise men". These systems are particularly dependent on the HTTP protocol.

Various common methods for load distribution are also state of the art, these being attributable to short-term operational load distributions. The utilization and availability of the respective possible provider systems are monitored by the technical systems "load balancers" and the orders are distributed accordingly and then executed. Long-term planned distributions are usually carried out using the "redirect" method based on corresponding HTTP requests, or on the basis of ff addresses. These long-term procedures do not cover the load distribution based on a dynamic DNS service.

Object of the invention

The object of the invention is to automatically determine the associated information address for requests for information (information requests). The technical system of the invention enables the integration of decentralized information in distributed networks regardless of the technology of the respective information system and the respective information provider. For this purpose, the system uses the existing standard protocols of the Internet to enable rapid dissemination and optimal use. For this reason, the existing and established system of the DNS protocol for determining the network addresses of computers is used and expanded for this information purpose, the technical internal functional sequence being exchanged in the sense of the invention. In addition, the technology is expanded in such a way that a dynamic fuzzy query to DNS servers can be processed depending on the context of the query and the dynamic resolution of the query to the network address of a corresponding "know-how" computer ("wise" ) he follows. This extension makes it possible to implement dynamic information systems based on distributed systems using the existing global DNS structure. This enables a new way of addressing and thus the distribution, search and provision of information. This represents a completely new technical access to the content of distributed systems, such as that of the Internet, which has some significant advantages over conventional addressing via URL offers. This technical process thus represents an optimal addition to the distributed content access. While URLs represent the address of a single piece of specific information (information atom), the invention of LUAR (Local Universe Address Resolution) addresses information units (suitable amount of information that consists of several URLs) can exist - also a hifi group). These information units are provided by the corresponding "wise men". The LUAR invention provides the addresses of these wise men. The knowledge management on which the dynamic address decisions are based is flexibly expandable and administrable. It supports the integration of various existing decision-making systems according to different criteria Existing systems from the market and your own can be combined to support personalization, localization, synonym handling, multilingualism, fault tolerance and targeted long-term load distribution. Furthermore, logical links are considered in the content information according to the respective technical status.

The system can be combined with known systems and methods of content search, content provision and load balancing and supports joint dependent and independent decision-making processes for the provision of information. In particular, LUAR can be combined with methods of "HTTP request" dependent personalization, regionalization, content search and provision as well as request redirection ("redirect") and load balancing and can be used in addition and in combination with existing Internet search engine technologies. The system can also be implemented as a much faster hardware solution and integrated into existing network components. This makes sense because the computational complexity of determining a dynamic information address is significantly higher than with current static DNS systems. technical solution of the task

The invention is implemented using and expanding the existing standardized DNS protocol and can thus be used directly in the first implementation stage instead of a conventional DNS server.

A DNS server of the last hierarchy level (which ultimately carries out the resolution) viewed the behavior from the outside as a "black box" (for the relevant case), responding to a request with a computer name of the type:

[<x-level-domain>.] .. [<third-level-domain>.] <second-level-domain>. <top-level-domain> to return the network address (IP address) of the computer assigned to this name (eg 192.168.241.15), which it determines from this entry in the "Domain" and "Zone-File" specifications. The solution of the invention in implementation variant 1 is achieved by changing the technological process between the request and the return value.

Instead of searching for the query exclusively from "domain files" and "zone files", a dynamic knowledge management system is queried, which determines a suitable know-how carrier for the query and the network address (IP address) of the computer of the " Sages "returns. This address is then returned to the client in question like any previous result of a query to a DNS server. In order for the existing information systems with their integrated DNS requests to function as before, the information request for the invention in implementation variant 1 must correspond to the syntax of a domain name. This results in the following query: [<x-th sub-term>.] .. [<main term>.] <second-level-domain>. <top-level-domain>

Example: Schnell.chinesische.vorspeisen.kochen.local-universe.com

The so-called "second-level" and "first-level" domain details are fixed and are internationally clearly assigned to natural and legal persons. In the technical process, these also determine the DNS server used for address resolution. At least two DNS servers are defined centrally when the respective domains are assigned. If the specified DNS server is LUAR-capable, according to the invention all further domain levels of the DNS request can be used as desired for information requests according to the rules of domain names. The knowledge management system involved can be implemented in a variety of ways. Just as certain information from the request is ignored in the simple case, any intelligent information system (knowledge base) can be used to find results. This is discussed in more detail in the example system.

As a result, the address of the information system (IP address) is returned, which can provide the information for the specific request or can provide qualified help. (This system, which is addressed by the address and provides the information, is referred to as a "wise man").

The qualitative leap from the determination of a static computer address to a dynamic information address takes place within the framework of the standardized DNS protocol. In implementation variant 2 of LUAR, the DNS protocol is expanded by some functions which are helpful for transparent personalization, regionalization and further flexibility, but which require the support of this extended protocol by the client-side applications, whose development is not in the hands of the Inventor lies. The extensions to the protocol correspond to two concrete changes. For one thing, the request with the domain name is supplemented by an additional piece of information, a property vector. The request is no longer just:

Give me the IP address for the following domain name:

[<x-level-domain>.] .. [<third-level-domain>.] <second-level-domain>. <top-level-domain> or

[<x-th sub-term>.] .. [<main term>.] <second-level-domain>. <top-level-domain> but a property vector is added to the query, each property with the associated values contains. This results in the request variant: Give me the IP address for the following domain name:

[<x-level-domain>.] .. [<third-level-domain>.] <second-level-domain>. <top-level-domain> taking into account the properties:

[<property> <separator property / value> <value>] ... [separator property> <propertyX> <separator property / value> <value>]

The property vector can e.g. can also be constructed according to XML syntax (see current XML standardization).

The second extension of the protocol optionally offers the return of a list of results (instead of the previous individual value) of addresses, with or without an indication of the weighting (hit accuracy) for each address of the return list. Support for this extended functionality by applications can only be expected if a specific version of this solution is defined as the standard.

Since implementation variant 2 can only be used to a relevant extent in the medium to long term (due to the need for standardization and application development on it), the extended functionality of personalization, regionalization and parts of the load and task distribution are combined using combinations of implementation variant 1 and Use of known "HTTP request" methods on information systems, the "wise men" realized. The following process follows. The LUAR system receives the request for information and uses the knowledge management system (Knowledge Base) to determine the currently associated address of the information system "Weisen", which contains this information. The application then uses HTTP requests with the same domain name This domain name, which represents the original information request, is now evaluated by the "wise man" plus all other parameters of an HTTP request and the possible further specializations of the information are determined therefrom. For this purpose, the respective information system "Wise Men" also has other specific information on its own system (or system group) available.

A short concrete example should illustrate this combination:

The request for the information "pizza.hawai-or-calzone.local-universe.com" is made by entering exactly this "domain name" as a URL in a WWW browser (eg Internet Explorer). The LUAR system, which at "local-universe.com" instead of a conventional one DNS server is used, receives the request for the address to "pizza.hawai.local-universe.com" from this WWW browser according to the DNS standard. The LUAR system determines from its knowledge base that it is an inquiry about "eating out" and the Italian food pizza, specifically referring to the two types of pizza "Hawaii" or "Calzone". The LUAR system uses this to determine the address of the relevant information service and returns it to the WWW browser. The WWW browser now sends its request "pizza.hawai.local- universe.com" to this address. Additional parameters are also supplied with an HTTP request, or are optionally and transparently supplied to the user. Assume that this is done If the system with the WWW browser is a GPS-capable system, the GPS position can be supplied directly in the HTTP header, for example from Berlin, and a user identification is also available in the HTTP request. The "manner" (information system ) breaks down the HTTP request and its parameters into the individual components and compiles the required information, "pizza.hawai.local-universe.com" prompts the wise men, for example, all Hawaii and Calzone pizza offers from Berlin (a specific region around the GPS -Position of the requesting client) If the user identification is known to the system, in the case of a subsequent order, for example from a pizza delivery service, the li eferadresse be determined automatically. This application example shows at least one possible implementation variant for HTTP services-based information services also in LUAR implementation stage 1 to emulate functional properties of LUAR implementation stage 2. In LUAR implementation level 2, personalization and localization can already take place in the LUAR and are therefore also available in an identical manner for services that are not based on HTTP requests (for example, TCP / IP services such as ftp, telnet, ssh, etc.). Another short example illustrates the application of the LUAR for planned, targeted and long-term load distribution in distributed information systems. The dynamic capabilities of the LUAR also allow the system to determine and return different addresses on the same request, based on the information in the Knowledge Base. Thus, if there are several systems for load distribution that can supply the same information, it does not matter to the user of information which system ultimately delivers the information. The LUAR Knowledge Base contains the information on these systems with the same information and their workload in the past and therefore provides a different, determined or randomly different one of the possible addresses after a correspondingly adapted procedure. According to the DNS standard, these addresses are retained in external client systems for up to 48 hours until they inquire about the current address at the LUAR again. As a result, the system can be used for a longer-term, specifically planned distribution of the system load on the information systems. This technology cannot react to short-term load peaks or the short-term failure of a system. However, it is ideally suited as a supplement to existing load balancing technologies, since the load balancing can be carried out across a wide range of networks without communication overhead and system dependencies.

Another important application variant of the invention is the combination of LUAR with any search engine technology that exists today. A search engine, for example for determining URLs on the Internet, is to be understood quite understandably as a “wise man” (information service of the distributed information network) in one or more information areas. If the LUAR determines this "wise man" for an information request, its address is returned. The request from an application, for example a WWW browser, is thus sent to the address of this wise man and in this special case to the search engine / search service. This request contains For example, in the case of HTTP requests, the specific information request with all the relevant terms as it was already available to the LUAR. In contrast to the DNS protocol, protocols such as HTTP provide additional information (e.g. all parameters of the HTTP request, such as user identification or position information ), which are used by the search engine to further qualify the content search. In this case, a search engine can immediately carry out the qualified search automatically and without the need for additional user input. This also makes sense to the extent that specialized subject-specific search services have more qualified knowledge to this Subject area than can be managed in the LUAR Knowledge Base. This two-step search is for special information, for example technical and scientific topics and keywords an effective and sensible variant. Two examples, a technical global and a local one, are intended to illustrate this application.

The request to the LUAR for "performance.apache-oder-squid.local-universe.com" returns the address of a "wise man" who is a specific search engine for "Internet Open Source Software". The WWW browser ( eg Internet Explorer) an HTTP request with "performance.apache-oder-squid.local-universe.com" to this address. The search engine can now e.g. automatically separate all terms before the "second-level domain" (everything before "local-universe.com") and automatically generate a search like "Performance AND (apache OR squid)". Without the user having to enter anything, he gets the search engine or the search service now displays the results list.

The "local" example shows the use of additional parameters of the HTTP request, whereby the example assumes that the GPS position of the requesting user can be seen from the HTTP request and that the user is currently in Berlin. The LUAR receives the request, for example "Mitarbeiter.sozialamt.local-universe.com" and determines the address of the information search service via the city administrations. This information search service determines the region "Stadt Berlin" from the GPS position and performs the search query automatically

"Infotype = employee, department = social welfare office, region = city, name = Berlin" and returns the list of employees of the social welfare office in Berlin. This example shows very well the complementary possibilities of information routing via this two-level search hierarchy and the possibilities and advantages for the users with these automatic information flows via several search technologies. Without LUAR and GPS localization, the above query is unthinkable on a global scale. Note: The same query can be used worldwide for the same local information (see LUAR properties) otherwise every city needs to know specific URLs, which is rather rare, especially when newcomers move, so that one request is sufficient worldwide. You do not have to know these in a specific language, correctly or with precisely defined terms. The contained separate representations of the sample application for the future information service Local Universe "describe more concrete detail solutions in detail.

Advantages of the invention

• The information search can be done directly using natural terms.

• The information search can be carried out immediately and directly, without first having to find an appropriate information service.

• The existing standard protocol of the DNS for determining computer addresses can be used in the invention for information address resolution. This means that the system can be used immediately without any technical changes on the user side (client side).

• The addresses of information from a wide range of information providers distributed worldwide can be determined in a simpler and more intuitive way than via URL (Uniform Resource Locator) (standard address of content on the Internet).

• The use of the invention enables the address determination of information which is transparent to the user. Therefore, the user can directly access the content. He is not first presented with lists from which he should determine himself what the correct result could be.

The addresses of the information can be determined dynamically, with different aspects such as Localization and system and network loads can be flexibly incorporated into the decision.

• Any combination of terms can be combined to describe the content.

• The system is largely language-independent. Information terms can thus be entered equally in different languages.

• It is also not possible to answer 100% correct queries (e.g. queries with spelling errors)

• Corresponding queries (synonym queries) can be answered. (e.g. via thesaurus functions) Client-specific information for determining the query result can be taken into account

(e.g. for personalized localized DNS - different results are assigned to the same query from Europe or America)

A multitude of different inquiries can be aggregated to a common result.

The information services ("wise men") themselves can use any technology in a variety of ways

Wisely and completely independently.

Various types of information services can be integrated into the system at any time without having to make changes to the respective information service itself.

Information services from any information provider can be integrated.

The basic LUAR system and the LUAR knowledge database can be used as both

Compact system or can be implemented as separate systems.

The knowledge database can also be implemented as a distributed system with several separate components. Systems from different systems can be connected to the interface of the knowledge database

Connect manufacturers.

It is no longer necessary to configure all information on the DNS server in the "Domain" and "Zone files", but can be developed and made available in one or more separate knowledge bases without having to reconfigure or re-configure the actual LUAR server to have to start.

Hiformation search systems can be set up based on the DNS protocol, which allows faster and more effective options than with conventional search engines based on URLs currently used. All search queries and dynamic computer access are bookmarkable and do not have to be entered in separate search query forms.

Possibility of integrating search queries for information with and without logical

Links already in the domain name.

Targeted long-term load balancing can be planned and carried out on the basis of dynamic DNS, which results in an optimal addition to existing load balancing systems.

The LUAR can be implemented both as a hardware system and as a software solution.

Results from knowledge base decisions are for quick processing and

Performance optimization cacheable.

In the case of software implementation, the effective solution can be applied via server processes capable of multithreading.

In the case of hardware implementation, the system can be integrated into corresponding network components.

The resulting results of the technical process of the content-dependent address determination enable the "wise men" (information nodes in the distributed network = system with the address determined by LUAR) to combine them with other HTTP request-based methods

Personalization, localization, load balancing, content search and content delivery.

An effective combination of the LUAR with conventional search engine technology is possible. The "wise" is then the corresponding information-specific search engine, which is based on the preselection (request to it) and the separation of the terms of the request in the

Domain names that can directly carry out subject-specific searches without the user having to enter or trigger anything else.

The knowledge database is not exclusively LUAR-specific and can therefore be used by others

Services (eg also the "wise men") are used. The information is therefore only to be maintained once, but can be used as desired ("edit once - use anywhere"). Example implementation of the patent application for:

Information search, information matching and information provision based on dynamic routing and address systems in networks -

Content-dependent dynamic DNS (Domain Name Service = Local Universe Address Resolution (LUAR)

Name of the example

"Dynamic Local Universe address resolution" - dynamic content-dependent DNS system for information search and information matching

introduction

The sample application is first shown in two graphical overviews and described in detail below. "Local Universe" is a future information network with local information and information access (including the search) via the URL (e.g. "italian.essen.local-universe.com"). For this purpose, the "dynamic Local Universe Address Resolution" was developed as an independent module for the DNS server "BIND" and this DNS server (open source software) was modified so that it uses this new functionality for address resolution.

description

The aim of the international localized information service is to reach the same content area worldwide at the same address. If I enter "pizza.local-universe.com" in Berlin, I get all pizza services in Berlin. If I enter the same URL in London, I get that from London. This means for scalable systems, depending on the location of the request another computer and consequently another IP address are returned. This functionality is implemented by "spatial search" (spatial search, position search) of the Knowledge Base. Often the user does not know the "correct" term that is expected to be entered. For example, he asks with the same wish for "italian.essen.local-universe.com". The same wish should lead to the same result and thus to the same information address. This is done by the synonym database (a kind of special thesaurus). Even a spelling mistake in the request does not change the user's wishes and should not lead to any errors (e.g. "pizzza.local-universe.com"). The fuzzy full-text search module is used for this.

As a result, content is not sought in the application of the "Local Universe Address Resolution" (LUAR), but rather content. The use of dynamic LUAR, together with the knowledge base, therefore forms a completely new type of search engine. Information is searched for using a URL and the request is forwarded transparently for the user to the computer, who knows about the requested information. The open source DNS server "BIND" was modified. All DNS requests to this modified DNS server are forwarded to the module "mod_luar", which establishes and maintains fixed connections to a daemon (server process - "luar daemon") in the knowledge base. So that already resolved queries can be answered quickly, a fast cache based on hash tables is also administered ("luar_cache"). The "luar daemon" is the server application that manages and executes the corresponding logic of the search queries in the knowledge database. The resulting IP address is then returned to the "mod_luar" (client) of the DNS server and the DNS server returns this result address to the requesting computer.

The knowledge base itself is managed and maintained via a web-based administration interface based on the "Java Server Page" technology. The WWW server is operated in SSL-encrypted mode, so that the knowledge database can also be maintained from outside the company's intranet can be done. This means that the knowledge base can be managed decentrally. The "Apache" WWW server and the JSP engine "Jakarta-Tomcat" are currently used as the WWW server and Java module. For performance reasons, all developments of the LUAR system itself ("mod_luar, luar-daemon, luar_cache") were implemented in the programming language "C" and translated with the "gcc" compiler under Linux.

The computer addresses (IP addresses) themselves are currently stored in the "Synonym Relations" area and can be flexibly expanded and assigned differently via the administration interface. This enables a dynamic expansion of the computer network managed by "know-how" carriers.

Since only LUAR of implementation variant 1 (standard DNS protocol) is transparently supported by the applications in the existing Internet, the functions of the "spatial search" module must first of all be carried out by the information services ("wise men") even using the HTTP request can be realized, which is disadvantageous for the local fault-tolerant scaling of the overall service, but will only be remedied with LUAR implementation variant 2.

Claims

Claims of the patent application for: Information search, information matching and information provision based on dynamic routing and address systems in networks - Content-dependent dynamic DNS (Domain Name Service) = Local Universe Address Resolution (LUAR) generic term / brief description Information search, information matching and information provision based on dynamic routing and address systems in networks Content-dependent dynamic DNS (Domain Name Service) = Local Universe Address Resolution (LUAR) is known from the current state of the art and the standard protocol of the Domain Name Service (DNS) is used based on "zone files" (configuration files) with defined (predefined) computer names or Domain names. This is the standard address resolution in TCP / IP networks, today's Internet. The behavior of the DNS server towards the requesting client is retained in exactly the same way. "Request in the syntax of a domain name" - "Return of an IP -Address of an associated computer ". In the subclaims, methods of localization, personalization and load distribution by the information services (" ways ") based on the current state of the art are integrated in the process combinations HTTP request. The knowledge base works for full text search , spatial search, for indexing and administration of information and knowledge using state-of-the-art methods. The implementation of the administration interface is also based on the current state of the art. Characteristic part (main claims / sub-claims)

1. Procedure and a technical. System for address resolution of information in distributed information systems, characterized in that this is done dynamically and content-specifically for an information request in order to enable dynamic automatic routing of information flows and targeted automatic access to information. An important part of the information request is already contained in the domain name.

2. The method according to claim 1, characterized in that the dynamic determination of the addresses of the information and information services takes place using and expanding the existing standardized DNS protocol.

3. The method according to one or more of claims 1-2, characterized in that the information search can be carried out directly using natural terms and in different languages.

4. The method according to one or more of claims 1-3, characterized in that the search for information can be carried out immediately and directly, without first having to seek out an appropriate information service.

5. The method according to one or more of claims 1-4, characterized in that the addresses of the information are determined dynamically, different aspects such as e.g. Localization and system and network loads can be flexibly incorporated into the decision.

6. The method according to one or more of claims 1-5, characterized in that any combination of terms can be combined to describe the content.

7. The method according to one or more of claims 1-6, characterized in that a plurality of different requests can be aggregated to a common result.

8. The method according to one or more of claims 1-7, characterized in that targeted load distributions are planned and carried out on the basis of the information designation via the information addresses.

9. The method according to one or more of claims 1-8, characterized in that an automatic information search based on the preselection and / or the separation of the terms in the domain name is carried out directly as a topic-specific search, without the user having to enter or trigger something additional ,

10. The method according to one or more of claims 1-9, characterized in that search engines of any technology can be integrated into the information search.

11. The method according to one or more of claims 1-10, characterized in that search requests for information with and without logical links are already specified in the domain name.