WO2009132713A1

WO2009132713A1 - Method and apparatus for assisted data extraction

Info

Publication number: WO2009132713A1
Application number: PCT/EP2008/060217
Authority: WO
Inventors: Carlo Martini
Original assignee: Carlo Martini
Priority date: 2008-04-30
Filing date: 2008-08-04
Publication date: 2009-11-05
Also published as: ITMI20080801A1

Abstract

A device (1) for assisted extraction of data from a data bank comprising: a first logic unit (2) for extracting and displaying a list of strings which contain an alphanumeric sequence which constitutes an input data item (10); a succession of second logic units (3, 4, 5, 6, 7) for extracting and displaying a list of strings which are associated with the input data item (10), provided in input to each logic unit (3, 4, 5, 6, 7), on the basis of the input data item (10) and of a selection in the string list, provided as an additional input to the logic unit (3, 4, 5, 6, 7) and originating from the preceding logic unit (2, 3, 4, 5, 6).

Description

METHOD AND APPARATUS FOR ASSISTED DATA EXTRACTION Technical Field

The present invention lies in the field of search engines, i.e., of tools that allow to retrieve information of interest by entering keywords. Background Art

The ubiquitous diffusion of digital technologies and the worldwide coverage provided by data communications networks currently make available an enormous amount of data containing information of several kinds, which can be used freely in several contexts. The currently best-known use of search engines is in the field of

Internet searches. Specifically provided engines, which can be accessed through known portals such as Google™ or Yahoo™, allow to locate Web pages whose content is relevant to keywords entered by a user.

However, search engines suffer substantial problems, which cause the search activity performed by the user to be tiring and scarcely efficient.

A first problem relates to the impossibility for the search engine to identify whether a given result is actually relevant to the keywords that the user has entered.

For example, by entering the word "Ferrari", the search engine is incapable of distinguishing among cars, cheeses or other items. Once the user has realized that he has to clarify better the subject of his search, he is forced to think up new keywords that might limit the results obtained. Having understood this, for example, the user performs a new search by adding the keyword "car" to the keyword "Ferrari". Once again, however, the search is inefficient, since the search engine cannot determine whether the user is interested in the results of the latest Formula One race or in searching for a car dealer.

All this, therefore, forces the user to a complex process of querying or navigating among the obtained results, repeating the search in each instance with different or increasingly detailed search strings in order to identify, among the countless generated results, the ones that are truly of interest to him.

Another problem that affects the background art relates to the size of the database and to the time required to query it. So-called index files are in fact used to optimize or reduce waiting times for the return of the results after sending a search string. However, these index files grow enormously when the information to be searched is heterogeneous in terms of type and substantial in terms of size. Disclosure of the Invention The aim of the present invention is to overcome the problems cited above with reference to the background art by devising a method and a system that allow to guide the user in extracting data that are indeed of interest to him in a simple and linear manner.

Within this aim, an object of the present invention is to allow a guided search in short times, both as regards the time used by the user during the search process and as regards the processor time required to perform physically the querying of the data bank.

Another object of the invention is to keep the size of the computer data required for the search within modest limits, reducing the need to use a large number of index files.

This aim and these and other objects, which will become better apparent hereinafter, are achieved by a method for assisted extraction of data from a data bank, which comprises the steps consisting of: entering an input data item, which is constituted by an alphanumeric sequence, as a search data item; displaying a list of strings that contain said alphanumeric sequence; making a selection from said list of strings; on the basis of said input data item and of said selection, displaying a list of strings associated with said data item; iterating the last two steps a number of times that corresponds to a number of assured logic bases, each of which corresponds to a different selection category. The intended aim and objects are also achieved by a device for assisted extraction of data from a data bank, which comprises: a first logic unit for extracting and displaying a list of strings that contain an alphanumeric sequence that constitutes an input data item; a succession of second logic units for extracting and displaying a list of strings associated with said input data item, provided in input to each logic unit, on the basis of said input data item and of a selection in said string list, provided as an additional input to said logic unit and originating from the preceding logic unit. Preferably, the logic units are six in number.

Conveniently, each string in the string list proposed by the extraction device in auto-completion can begin with said alphanumeric sequence or can contain said alphanumeric sequence.

Moreover, the alphanumeric sequence can comprise internally a separator, for example a space or a dash, which divides the sequence into two or more substrings, from which it is possible to extract a more significant substring.

The iteration in the extraction process can therefore be performed by using as input for each logic unit the entire alphanumeric sequence and/or the most significant substring.

Brief Description of the Drawings

Further characteristics and advantages of the invention will become better apparent from the following detailed description, given by way of non-limiting example and accompanied by the corresponding figures, wherein:

Figure 1 is a first view of the device according to the invention; Figure 2 is a schematic view of the device according to the invention, provided in TTL logic;

Figures 3 to 6 exemplify the steps of a data extraction process performed by applying the teachings of the present invention. Ways of carrying out the Invention

Figure 1 is a view of a preferred embodiment of the system according to the invention, in which six logic units are used.

In particular, Figure 1 illustrates schematically a device 1 according to the invention, which comprises a first logic unit 2 and second logic units 3, 4, 5, 6, 7.

Each logic unit has a first input, which is adapted to receive an input or search data item 10, which originates from data entry means 8, the input data item 10 being optionally processed in order to extract from it a more significant portion and being multiplexed by a multiplexer 9.

Each of the second logic units 3-7 further has a second input, which is adapted to receive the output of the directly preceding logic unit 2-6.

Finally, each logic unit has a first communication channel 12-17 for accessing a respective pattern 31, 32, 33 of a database 30 and a second communication channel which is adapted to delimit the application of the respective pattern 31, 32, 33 on the basis of subsets 22-21.

Operation of the system according to the invention is as follows.

The user accesses a search window, which is connected to the data bank 30, and starts to enter in appropriate entry means 8, for example a text box, an alphanumeric sequence that corresponds to a search data item 10.

The data item 10 is passed in input to the first logic unit 2, which checks within the pattern 31 of the database 30 for the presence of terms that begin or alternatively comprise the sequence of characters that the user is entering, suggesting completion of the sequence on the basis of the retrieved data.

Once the user has finished entering the alphanumeric sequence, if necessary by selecting it directly among the options proposed by the logic unit 2, the input data item is passed in input to the multiplexer 9, which replicates the data item in order to provide it in input to the second logic units 3-7. If the multiplexer 9 identifies, within the alphanumeric sequence, the presence of a separator, for example a space, a comma, a dash or others, it may identify a more significant substring before passing the data item to the logic units, as will become better apparent hereinafter. At this point, the device 1 triggers a mechanism of sequential selection among the generated possible results, in which each logic unit provides in output an assured result that belongs to the corresponding subset, allowing the user to proceed by successive steps in identifying the desired result or results. By working in this manner, the user's progress in the search is always kept under control, since it is constrained by the existence of assured data and is performed rapidly.

For the sake of clarity in explanation, and merely by way of non- limiting example, an example of application of the invention is now described with reference to Figures 3-6 in a context of a data bank related to commercial organizations located geographically.

Starting from the numeric calculation of the assured company codes that look onto: roads, streets, squares, alleys and so forth, it is possible to measure, at preset thresholds, whether the region being analyzed has predominantly one of the following main uses: mercantile, commercial, handicraft, agricultural or industrial.

By means of the functionality of a simple vector algorithm, which can be defined easily by the person skilled in the art and is relevant to the maximum assured value that can be expressed and is related to companies that are present along the geometric extension being considered, and to the length, in polar coordinates, of the geographic element being considered, it is possible to orient a logic decider with extreme precision.

For example, considering an indicator scale from 0 to 9 inclusive, if, for a given road, street or square, one considers 20 assured activity codes, of which 3 are for handicraft expression, 5 are for mercantile expression, 8 are for commercial expression and 4 are for industrial expression, it is possible to orient the logic decider, i.e., a module which selects whether the input can be included in preset categories or selections by being based on a decreasing statistical order, which refers to the maximum number of obtained homogeneous codes. In this regard, the oriented output corresponds to a previous homogeneous assured data item, in order to obtain, by means of a basic logic "and", a new outcome, for selection, which provides a further indication and is even more oriented toward a decision model which is as error-free as possible. With reference to Figure 3 and to the logic unit 2, the user enters the data item to be searched, "Mantova" in the example. Progressive auto- completion leads the user to enter correctly the search data item and therefore refer to an assured data item.

With reference to Figure 4 and to the logic unit 3, the user selects the the "district" in which the "locality" is present. In this case also, progressive auto-completion leads to the selection of an assured data item, which is identified in the example with "Province of Mantova" (Provincia di Mantova in Figure 4).

With reference to Figure 5 and to the logic unit 4, the user makes a selection among the municipalities that are present in the province of Mantova, for example "Mantova" itself, again by utilizing the progressive auto-completion to select an assured data item.

After selecting, to continue, a subunit for selection between "Tourism" and "Shopping", with reference to Figure 6 and to the logic unit 5, the user enters the name of the Street, Square, Alley, Road or others, for example "Erbe Square" (Piazza Erbe in Figure 6).

With reference again to Figure 6 and to the logic unit 6, the user selects a street number and, if present, the suite, stairway, courtyard, floor, and section. Finally, again with reference to Figure 6 and to the logic unit 7, the user selects the commercial concern located within a single Shopping Center.

By working in this manner, the device 1 according to the invention has allowed the user to perform an assisted extraction rapidly and effectively, unambiguously oriented toward the required result.

Substantially, considering again the preceding example, and to better clarify the concepts described above, the expression "assured logic base" is used to reference a homogeneous or type-based classification of the data of a database. For example, one logic base can be "districts" and contains all the words that indicate a district.

For example, a subsequent logic base can be "municipalities" and contains all the words that indicate a municipality, each linked to a district.

A subsequent logic base can be "roads" and comprises all roads, each linked to a municipality. A subsequent logic base can be "shops" and comprises all shops, each linked to a road.

The "shops" logic base that comprises all shops can also be linked to other logic bases, for example directly to "municipalities".

The method provides for guided data extraction, in which, since a selection is iterated within "assured" data bases (it is not possible to search for a word that does not exist, if one searches for "Ferrari", depending on the logic base being used, one knows whether the Ferrari car or the Ferrari salami will be output), one obtains a result which necessarily matches the subject of the search, making the search or selection much faster. In practice it has been shown that the present method and device achieve the proposed aim and objects. In particular, it has been shown that the described method and device allow to obtain, by exclusive logic extraction, assured results, within a mass of assured data equal to l * 10^Λ6

(1,000,000) which are homogeneous one another in terms of origin, comparable in terms of purpose, semantically relevant in terms of objectivity and standardized in terms of spelling. In this regard, the requesting data item, entered in the linear extraction matrix, is multiplexed in a base-six nonintersectable (identical to itself) mode, compared for homogeneity by six different parallel inputs, which originate from six different bases of assured data; compared as regards relevance by six different inputs which are semantically relevant to a finite data set and subsequently compared again, by means of a literal subdivision, with six inputs which in turn are assured, in order to finally obtain an assured data item to be presented. Clearly, numerous modifications are evident and can be performed promptly by the person skilled in the art without abandoning the scope of protection of the appended claims. For example, it is obvious for the person skilled in the art that the invention can be provided in manners which are more oriented toward software implementation or toward hardware implementation, with logic circuits as shown schematically in Figure 2.

It is also evident that the inventive concept on which the present invention is based is independent of the actual implementation of the software modules, which can be provided in any language and on any hardware platform, and that the method of integration with the different information technology systems, Web browsers and others, which can utilize the invention to achieve extraction of the desired assured data, is also non-limiting for the purposes of the invention.

Therefore, the scope of protection of the appended claims must not be limited by the illustrations or by the preferred embodiments illustrated in the description by way of examples, but rather the claims must comprise all the features of patentable novelty that reside in the present invention, including all the characteristics that would be treated as equivalent by the person skilled in the art. The disclosures in Italian Patent Application no. MI2008A000801, from which this application claims priority, are incorporated herein by reference.

Where technical features mentioned in any claim are followed by reference signs, those reference signs have been included for the sole purpose of increasing the intelligibility of the claims and accordingly such reference signs do not have any limiting effect on the interpretation of each element identified by way of example by such reference signs.

Claims

1. A method for assisted extraction of data from a data bank, comprising the steps of: a) entering an input data item (10), which is constituted by an alphanumeric sequence; b) displaying a list of strings that contain said alphanumeric sequence; c) making a selection from said list of strings; d) on the basis of said input data item (10) and of said selection, displaying a list of strings associated with said data item (10); e) iterating steps c) and d) a number of times that corresponds to a number of assured logic bases, each of which corresponds to a different selection category.

2. The method for assisted data extraction according to claim 1, characterized in that each string in said list of strings begins with said alphanumeric sequence.

3. The method for assisted data extraction according to claim 1 or 2, characterized in that said number of assured logic bases is six.

4. The method for assisted data extraction according to one or more of the preceding claims, characterized in that said logic bases comprise one or more among: locality, district, municipalities, address, street number, detail of street number.

5. The method for assisted data extraction according to claim 4, characterized in that said alphanumeric sequence comprises a separator which divides said sequence into two or more substrings.

6. The method for assisted data extraction according to claim 5, comprising the step of identifying, in said two or more substrings, a more significant substring.

7. The method for assisted data extraction according to claim 6, characterized in that the iteration of said steps c) and d) is performed by using as input data item said entire alphanumeric sequence.

8. The method for assisted data extraction according to claim 6, characterized in that the iteration of said steps c) and d) is performed by using as input data item said more significant substring.

9. A device (1) for assisted extraction of data from a data bank, comprising:

- a first logic unit (2) for extracting and displaying a list of strings which contain an alphanumeric sequence which constitutes an input data item (10); - a succession of second logic units (3, 4, 5, 6, 7) for extracting and displaying a list of strings which are associated with said input data item (10), provided in input to each logic unit (3, 4, 5, 6, 7), on the basis of said input data item (10) and of a selection in said string list, provided as an additional input to said logic unit (3, 4, 5, 6, 7) and originating from the preceding logic unit (2, 3, 4, 5, 6).

10. The device (1 ) for assisted data extraction according to claim 9, characterized in that each string in said list of strings begins with said alphanumeric sequence.

1 1. The device (1) for assisted data extraction according to claim 9 or 10, characterized in that said succession of second logic units (3, 4, 5, 6, 7) comprises five logic units.

12. The device (1 ) for assisted data extraction according to one or more of claims 9 to 11, characterized in that each of said logic units (2, 3, 4, 5, 6, 7) is connected to a different logic base.

13. The device (1) for assisted data extraction according to claim 12, characterized in that said logic bases comprise one or more among: locality, district, municipalities, address, street number, street number detail.

14. The device (1) for assisted data extraction according to claim 13, characterized in that said alphanumeric sequence comprises a separator which divides said sequence into two or more substrings.

15. The device (1) for assisted data extraction according to claim 14, characterized in that one or more of said logic units (2, 3, 4, 5, 6, 7) comprise means for identifying, within said two or more substrings, a more significant substring.