DE19843450A1 - Electronic Thesaurus for internet search machine - Google Patents

Abstract

The electronic Thesaurus contains lists which each have at least one main concept and at least one subsidiary concept which is associated with the main concept, each data information scanned to extract the subsidiary concepts, with the corresponding main concepts identified. A concept vector is provided for each electronic data information using vector components provided by the different identified main concepts, with the structural relationships between the electronic data information provided by the concept vectors.

Description

The invention relates to the determination of a structure of a Variety of electronic files in terms of an electroni Thesaurus.

From [1] it is known the contents of electronic files compare with each other and a structure of the electroni represent a content-related file Similarity of the files in the form of a spatial proximity in ei is described in an information space.

Under an electronic file is, for example, a to understand electronic text file.

Under an electronic thesaurus is in a further ge arranged electronic dictionary of terms and theirs Designations (References) to a predefinable strukturier th subject area.

An example of a thesaurus for the field of medicine is known from [2].

The terms are in the electronic thesaurus from [2] in electronic lists broken down, each list one Generic term is assigned, and each list is a predefinable Number of sub-terms has, which also the jewei associated with the generic term. The sub-concepts are For example, synonyms for the generic term or Be handles that are characteristic of the generic term.

A so-called Internet search engine is known from [3].  

The invention is based on the problem, a method and an arrangement for determining a structure of a plurality electronic files regarding an electronic The saurus in which or for which the structure has a ver improved reliability in terms of comparability of electronic files with each other is achieved.

The problem is solved by the method according to claim 1 and solved by the arrangement according to claim 10.

In the method for computer-aided determination of a Structure of a variety of electronic files regarding of an electronic thesaurus contains the electronic The saurus electronic lists, one list each one Generic term is assigned and wherein a list each min at least has a sub-concept to the preamble to is ordered. In every electronic file every sub is term of the electronic thesaurus determines which in the file is included. For every electronic file is each subclause is mapped to the generic term, the Sub-term is assigned and for each electronic file a generic term vector is determined in which at least one Specification for at least part of the generic terms included is whether a generic term of this part in the respective elek tronic file is included. From the generic terms the structure between the electronic files is obtained telt.

The arrangement has a processor which is set up in such a way that the following steps are feasible:

• a) the electronic thesaurus contains electronic lists, wherein a list is assigned to a generic term and wherein a list is at least one sub-term which is assigned to the preamble,
• b) in each electronic file, each subclause of the electronic thesaurus determines which in the file is included  
• c) each sub-term is deducted for each electronic file forms on the generic term, to which the subordinate concept is assigned is net,
• d) every electronic file becomes a generic term vector determined, in which at least one statement for at least one Part of the generic terms is included, whether a generic term this part in the respective electronic file hold, and
• e) the structure is interposed between the generic vectors the electronic files.

By the invention it is achieved that each electronic Da Very good on the basis of the generic terms with the others electronic files can be compared. In this Wei se is an improved comparability and thus a ver improved searchability of electronic files enabled light.

Preferred developments of the invention will become apparent from the dependent claims.

It is provided in a development of the invention that the generic term vector in each case the number one in the jewei has a single electronic file contained in the generic term.

In this way is also the frequency of occurring Generic terms in an electronic document in the context of Structural analysis takes into account what is left of you results in a more comparable result.

The invention is very advantageous in a distributed re network applicable, whereby the electronic files on ver different computers connected to each other via a computer network are stored. The computer network can at spielsweisle be the Internet.  

In this case, the electronic files of the ver different computers determined based on a given Be user request. The electronic files are read and stored in the computer and then according to the invention further processed. Thus, a very reliable automati based search tool in a distributed computer network ge create.

It is provided in a further embodiment that Be Start the process in every electronic file every word that is not deleted in the electronic thesaurus hold is. In this way, the feasibility of the Process considerably accelerated.

The invention is in any technical field can be used, d. H. the electronic thesaurus can be on a be lovely definable area, for example medicine, spe be cialized.

An embodiment of the invention is shown in the figures and will be explained in more detail below.

Show it

Fig. 1 is a block diagram showing an embodiment of the invention;

Fig. 2 is a sketch of a computer network with a plurality of computers.

Fig. 2 shows in a sketch a plurality of computers Ri (i = 2... N) and a first calculator R1 nernetz a computing RN are connected together. The computer network RN is a computer network RN, which carries out the transmission of digital data in accordance with the TCP / IP protocol (Transport Control Protocol / Internet Protocol).

A user submits a user request to the first computer R1 on, from the first computer R1 as a query A on the plurality of computers Ri is transmitted.

In each computer Ri are elek according to known search tronic files Dj (j = 1 to m) determines their content with regard to the request A is more similar than a predefinable first threshold.

Such a determination shall be made using a standard Internet Search engine described in [3].

Each computer Ri sends in an answer PDi the electro nical files Dj, which Ri ermit from the respective computer telt, to the first computer R1.

From the first computer R1 all electronic files Dj are read in via an input / output interface 201 and stored in a memory 202 , which is connected to the input / output interface 201 via a bus 203 . In the memory 202 , an electronic The saurus 204 is further stored.

In the following it is assumed that the electronic thesaurus 204 six electronic lists L1, L2, L3, L4, L5 and L6 has.

Each electronic list has a generic term OB1, OB2, OB3, OB4, OB5, OB6. Furthermore, each electronic list at least one subclause. Each subclause of a electronic list is the respective generic term OB1, OB2, OB3, OB4, OB5, OB6 assigned.

In the following, the following set of electronic lists L1, L2, L3, L4, L5 and L6 is assumed as the content of the electronic thesaurus 204 :

First list L1 First generic term OB1: AIDS First sub-term UB11: HIV-1 Second sub-term UB21: HIV-2 Third sub-term UB31: SIV

Second list L2 Second generic term OB2: pathogen First sub-term UB12: virus Second sub-term UB22: bacterium Third sub-term UB32: microbe Fourth sub-term UB42: bacillus

Third list L3 Third generic term OB3: smallpox First sub-term UB13: chickenpox Second sub-term UB23: variola Third sub-term UB33: Scroll

Fourth list L4 Fourth generic term OB4: drug First sub-term UB14: drug Second sub-term UB24: medicine Third sub-term UB34: Pharmacon Fourth sub-term UB44: preparation Fifth sub-term UB54: medicine Sixth sub-term UB64: remedy Seventh sub-term UB74: medium Eighth sub-term UB84: Parmaceutikum

Fifth list L5 Fifth generic term OB5: doctors First sub-term UB15: doctor Second sub-term UB25: doctor Third sub-term UB35: doctor Fourth sub-term UB45: healer Fifth sub-term UB55: therapist Sixth sub-term UB65: quack

Sixth list L6 Sixth generic term OB6: hospital First sub-term UB16: clinic Second sub-term UB26: hospital Third sub-term UB36: sanatorium Fourth sub-term UB46: hospital Fifth sub-term UB56: sanatorium Sixth sub-term UB66: Hospitals Seventh sub-term UB76: Hospital.

The method described below is performed in a Prozes sensor 205 , which is also connected via the bus 203 to the input / output interface 201 and the memory 202 of the first computer R1, the following steps performed by (see Fig. 1).

The electronic files Dj are read out of the memory 202 in a first procedural step (step 101 ) and all the words in each electronic file Dj are deleted which are not contained in the electronic thesaurus 204 as a generic term and / or sub-term.

Thus, m intermediate files ZDj are formed. Each intermediate file ZDj contains only words contained in the electronic thesaurus 204 .

In a further method step (step 102 ), for each intermediate file ZDj each sub-concept is mapped to the generic term of the electronic list to which the respective sub-term is assigned.

In a further step, multiply occurring Oberbe Errors for an intermediate file ZDj deleted, whereby Oberbe handle vectors OBVj (a generic term OBVj is ever because a file for an intermediate file ZDj) are formed.

In each generic term vector OBVj is thus a list of Contain a generic term, to each of which at least one Un term term in the associated intermediate file ZDj, d. H. also the corresponding file Dj.

In a third step (step 103 ), the similarity of the upper-level vectors OBVj with one another is determined from the generic term vectors OBVj. This is done by pairwise comparison of the generic term vectors OBVj and one after another according to the method known from [1].

Thus, a structure arises between the plurality of electronic files Dj, which is displayed to the user on a screen 206 connected to the first computer R1.

The representation is made such that content similar upper term vectors OBVj cause the corresponding Da tej symbolically spatially close to each other and close to the dj respective generic term are shown. Details of the Vorgehehsweise to transform the corresponding Vorge method of determining the spatial representation of the ancestors The nature of the files Dj can be seen in [1].  

An alternative to the embodiment described above is shown below:
It is possible that each generic term vector OBVj does not contain the generic terms themselves, but rather a number which describes the occurrence frequency of the respective generic term in the electronic file.

This document cites the following publications:
[1] T. Führing et al. Contextual Design: A Metaphor for the Visualization of and Interaction with Complex Knowledge, W. Rauch et al (Editor) Added Value of Information - Professionalization of Information Work, Proc. of the fourth intern. Symp. For Information Science, ISI 1994,
[2] DIN standard 1463
[3] O. Kyas, Internet Professional - Technological Fundamentals & Practical Usage, ISBN 3-8266-4002-0, International Thomson Publishing, Bonn, pp. 513-523, 1996

Claims (18)

1. A method for computer-aided determination of a structure of a plurality of electronic files with respect to an electronic thesaurus,

• a) in which the electronic thesaurus contains electronic lists, wherein a list is assigned to each a generic term and wherein a list each has at least one sub-term, which is assigned to the preamble,
• b) in which each sub-term of the electronic thesaurus is found in each electronic file, which is contained in the file,
• c) in which, for each electronic file, each sub-term is mapped to the generic term to which the sub-term is assigned,
• d) in which for each electronic file a Oberbegriffsvek gate is determined, in which at least an indication of at least a part of the generic terms is included, whether a generic term of this part is contained in the respective electronic file, and
• e) in which the structure between the electronic files is determined from the generic vectors.

2. The method according to claim 1, in the at least part of the lists several sub-concepts contains. 3. The method according to claim 1 or 2, in which the generic term vector in each case the number of an in the generic term contained in the respective electronic file having. 4. The method according to any one of claims 1 to 3, where at least a part of the subclauses synonyms of Represent generic term.   5. The method according to any one of claims 1 to 4,

• 1. in which the electronic files are determined by different, interconnected via a computer network computers due to a given user request, and
• 2. in which the electronic files are read in and stored.

6. The method according to any one of claims 1 to 5, every word deleted in every electronic file is not included in the electronic thesaurus is. 7. The method according to any one of claims 1 to 6, where the electronic thesaurus only terms of a pre technical area. 8. The method according to claim 7, where the technical field is medicine. 9. The method according to any one of claims 5 to 8, where the computer network is the Internet. 10. Arrangement for determining a structure of a plurality of electronic files with respect to an electronic The saurus,
with a processor arranged such that the following steps are feasible:

• a) the electronic thesaurus contains electronic lists, a list being respectively assigned to a generic term and a list each having at least one sub-term associated with the generic term,
• b) in each electronic file each sub-term of the electronic thesaurus is determined, which is contained in the file,
• c) for each electronic file, each sub-term is derived from the generic term to which the sub-term is assigned,
• d) for each electronic file, a generic term vector is determined, in which at least an indication is included for at least a part of the generic terms, whether a generic term of this part is contained in the respective electronic file, and
• e) the structure between the electronic files is determined from the generic vectors.

11. Arrangement according to claim 10, wherein the processor is arranged such that at least some of the lists contain several sub-terms. 12. Arrangement according to claim 10 or 11, in which the processor is arranged such that the upper term vector in each case the number one in the respective has electronic file contained generic term. 13. Arrangement according to one of claims 10 to 12, wherein the processor is arranged such that at least a part of the sub-terms synonyms of the generic term darstel len. 14. Arrangement according to one of claims 10 to 13, with a plurality of computers, in each of which at least one elec tronic file is stored, wherein the computer via a computer network with each other,
in which the processor is arranged such that

• 1. the electronic files can be determined by the computers on the basis of a given user request, and
• 2. in which the electronic files are read in and stored.

15. Arrangement according to one of claims 10 to 14, wherein the processor is arranged such that in each electronic file is deleted every word that is not in contained in the electronic thesaurus.   16. Arrangement according to one of claims 10 to 15, in which the processor is set up so that the elec tronic thesaurus only terms of a given techni area. 17. Arrangement according to claim 16, where the technical field is medicine. 18. Arrangement according to one of claims 14 to 17, where the computer network is the Internet.