JPH03122768A - Indexing supporting system - Google Patents

Abstract

PURPOSE:To improve the reproducing ratio of retrieval by providing the indexing supporting device with a thesaurus holding the relating information of plural index words and a relating index word retrieving part for retrieving a relating index word for an extracted index word in a reference. CONSTITUTION:The relating index word retrieved from the thesaurus 5 by the relating index word retrieving part 6 is also displayed and selected together with an extracted index word extracted from the reference to be registered at the time of selecting the index word by a result displaying/selecting part 3. Even in the case of a phrase other than the ones included in the reference, the relating phrase is properly applied to the reference as an index word to form a data base. Since a user can more properly expressed the concept of the reference, the selecting margin of index words can be expanded at the time of retrieving the reference and the reproducing ratio of retrieval can be improved.

Description

[Detailed Description of the Invention] Industrial Application Field The present invention allows a user to select an index word that is considered appropriate from among index word candidates extracted from documents to be registered in a database. The indexing that registers the documents in the database is related to the support device.

Conventional technology In general, this type of automatic indexing research is based on assigning a word string to each document as an index term that well expresses the content of a document or document set and that can sufficiently identify documents or document sets. It is intended to.

For example, the literature ■ ``Automatic indexing is a research trend'' (Information Processing Society of Japan Journal, Vol. 25, 9, 1984) and the literature ■ ``
"Automatic extraction of important words in Japanese literature" (Information Processing Society of Japan Journal, Vol. l 7. N11L2. l 9
76), IBM's STA I R8
, U.S. DDC mechanical auxiliary index, JICST JAKAS
, Kyoto University's SMART system, etc. are systems that extract words from documents, apply unnecessary word removal, grammatical rules, etc., and present several index word candidates to the user.

Figure 2 shows the system configuration of such a conventional automatic index extraction device.
automatically extracts index word candidates (extracted index words), displays the results to the user in the result display/selection section 3, and selects the candidate that seems appropriate from among the candidates displayed to the registration operator. By selecting index words, each document is indexed and an index file 4 is created as a database.

Problems to be Solved by the Invention However, in these conventional systems, the only index terms that can be assigned to a document to be registered are words that appear in the document. Strictly speaking, spellings that are not the same as the index word candidates in the document may be assigned due to variations in spelling or synonym processing, but these are recognized and registered as the same word, so essentially There is no change in the fact that only the words and phrases in the documents are added as indexes.

Due to the assignment of such limited index terms, these index terms must be entered strictly during the search, which is one reason why the recall rate of the search is low or the search is likely to be overlooked frequently. Become.

Means for Solving the Problems The invention according to claim 1 provides a thesaurus that holds index terms and relationship information between each index term, and relationships related to extracted index terms extracted from documents to be registered in a database. It consists of a related index word search section that searches for index words from the thesaurus, and a result display/selection section. The index word selected by the result display/selection section is configured to be stored in the database together with the document to be registered.

In the invention set forth in claim 2, instead of the related index word search unit of the invention set forth in claim 1, the thesaurus is searched for related index words related to extracted index words extracted from documents to be registered in the database. and a related index term search unit that calculates the importance of this related index term with respect to the document using a predetermined calculation formula with reference to the thesaurus and creates a list of related index terms in order of relative speed, and extracts the extracted index term. and the created list of related index terms in order of relative speed are displayed in the result display/selection section, and the index terms selected by the result display/selection section are stored in the database together with the document to be registered. It was configured to be stored.

When an index word is selected by the action result display/selection section, the related index word searched from the thesaurus by the related index word search section is displayed along with the extracted index word extracted from the document to be registered for selection. Therefore, even if the words and phrases are outside the content of a document, related words can be appropriately assigned to the document as index terms and compiled into a database. Therefore, the user can more appropriately express the concept of the document, and the room for selecting index words when searching for documents is expanded, and the recall rate of the search is improved.

In particular, by calculating the importance of related index terms with reference to a thesaurus, creating a list of related index terms in order of relative speed, and displaying the results along with a list of extracted index terms for selection.
It is possible to register all related index terms that are highly related, and more appropriate index terms can be assigned. Therefore, it becomes possible to search for documents that would otherwise have been missed due to incomplete search conditions in the conventional method.

Example An embodiment of the present invention will be described based on FIG.

Components that are the same as those shown in FIG. 2 are designated by the same reference numerals, and explanations thereof will be omitted. In this embodiment, in addition to the system configuration shown in FIG. 2, a thesaurus 5 is provided which holds all existing index terms as well as relationship information between each index term. Here, the relationship information between index words includes, for example, hypernym/hyponym/
This is information about related words, etc. In addition, the index automatic extraction unit 2
A related index word extraction unit 6 is provided which searches for related index words by referring to the thesaurus 5 for index word candidates (extracted index words) obtained by analyzing a document file.

In such a configuration, when a document to be registered is input to the automatic index extraction section 2, morphological analysis is performed and each sentence is divided into words. For these word groups, removal of spelling variations, conversion of synonyms to unified notation, and removal of unnecessary words are performed to generate index word candidates (extracted index words). Next, the related index word search unit 6 refers to the thesaurus 5 and searches for a related index word related to the extracted index word. After the search, the importance of these related index terms for the document to be registered is calculated using a predetermined formula.

The degree of importance can be determined, for example, using the following formula.

First, the importance takes a value between [0, 1], where 0 indicates unrelated and 1 indicates most related.

Second, the importance of the index word extracted from the document d to be registered is assumed to be l. Furthermore, the importance level Rd(i) of an index word i that does not exist in this document d is calculated as follows: N is the total number of index words extracted from document d, and the importance level Rd(i) of index word i that does not exist in this document d is determined by If the set is Ad(i), the frequency of appearance of index word j in document d is Nd(j), and the total frequency of appearance of index word in document d is Na1l, then Rd(i) = Σ N d (j) / Na1lJε
It is determined by Ad(i). Furthermore, the frequency of appearance in a virtual document can be defined depending on the difference in related ministries (superior/lower/related, etc.). For example, the appearance frequency Nd(j) in a document is 0,9 for a superordinate relationship, 0.3 for a subordinate relationship, and 1.0 for a related word relationship.
Substitute the value N'd(j) multiplied by
' Just substitute all.

After such importance calculation, a list of related index words in order of relative speed is created, and the result display/selection unit 3 displays this list of related index words in order of relative speed to the user along with a list of extracted index words. , offered for selection.

Effects of the Invention As described above, the present invention includes a thesaurus that holds related information between index words and a related index word search section that searches for related index words for extracted index words in documents. When the selection unit selects an index term, related index terms are displayed together with the extracted index term for selection, so even if the words are not in the document, related terms can be added to the document as an index term as appropriate. It can be made into a database,
Therefore, the user can more appropriately express the concept of the document, expand the scope for selecting index words when searching for documents, and improve the recall rate of the search. According to the invention, the importance of related index terms is calculated with reference to a thesaurus, a list of related index terms in order of relative speed is created, and the results are displayed together with a list of extracted index terms for selection. It is possible to register highly related index terms without omission, and to assign more appropriate index terms, thereby making it possible to search for documents that would otherwise have been missed due to incomplete search conditions in the conventional method. I can do it.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention, and FIG. 2 is a block diagram showing a conventional example.

Claims (1)

[Claims] 1. A thesaurus that holds index terms and relational information between each index term, and a search from the thesaurus for related index terms related to extracted index terms extracted from documents to be registered in the database. The extracted index words and searched related index words are displayed in the result display/selection part and selected by the result display/selection part. 1. An indexing support device, characterized in that said index words are stored in said database together with said documents to be registered. 2. A thesaurus that holds index terms and relational information between each index term, and a search for related index terms related to the extracted index terms extracted from the documents to be registered in the database from the thesaurus and a predetermined calculation formula. The system includes a related index word search unit that calculates the importance of this related index term with respect to the document with reference to the thesaurus and creates a list of related index terms in order of relative speed, and a result display/selection unit. The list of extracted index terms and the created list of related index terms in order of relative speed are displayed in the result display/selection unit, and the index terms selected by the result display/selection unit are registered. An indexing support device characterized in that the indexing support device is configured to store the same information in the database.