CN-122019747-A - Enterprise name retrieval method and device, electronic equipment and storage medium

CN122019747ACN 122019747 ACN122019747 ACN 122019747ACN-122019747-A

Abstract

The embodiment of the invention provides an enterprise name retrieval method, an enterprise name retrieval device, electronic equipment and a storage medium, wherein the method comprises the steps of receiving enterprise query text input by a user and obtaining corresponding structured semantic data; the method comprises the steps of obtaining an enterprise knowledge graph, inquiring a related candidate enterprise set, inquiring a first target enterprise consistent with enterprise inquiry text in the candidate enterprise set and giving scores, identifying a second target enterprise containing character strings of the enterprise inquiry text in the candidate enterprise set divided by the remaining enterprises of the first target enterprise and giving scores, inquiring a third target enterprise with enterprise names matched with the enterprise inquiry text semantically in the candidate enterprise set divided by the remaining enterprises of the first target enterprise and the second target enterprise and giving scores, and sorting according to the scores of all enterprises in the candidate enterprise set to generate a retrieval result list. The search method and the search device can better understand the query intention of the user, and greatly improve the search accuracy and the user experience.

Inventors

LEI XINGBANG
LIU JIN
XIAO QIYUAN
Ju Huayang
ZHUANG YUFEI

Assignees

中国长江三峡集团有限公司

Dates

Publication Date: 20260512
Application Date: 20260106

Claims (11)

1. A business name retrieval method, the method comprising: receiving enterprise query text input by a user, and acquiring structural semantic data corresponding to the enterprise query text, wherein the structural semantic data comprises a core name field; acquiring an enterprise knowledge graph, and inquiring a candidate enterprise set associated with the enterprise inquiry text in the enterprise knowledge graph based on the core name field; Querying a first target enterprise consistent with the enterprise query text in the candidate enterprise set, and assigning a score to the first target enterprise; identifying a second target enterprise containing a character string of the enterprise query text in the remaining enterprises of the candidate enterprise set except the first target enterprise, and assigning a score to the second target enterprise; querying a third target enterprise with enterprise names semantically matched with the enterprise query text in the rest enterprises of the candidate enterprise set except the first target enterprise and the second target enterprise, and assigning scores to the third target enterprise; And sorting according to the scores of all enterprises in the candidate enterprise set, and generating a retrieval result list according to the sorting result.
2. The method for searching for business names according to claim 1, wherein the ranking according to the scores of the businesses in the candidate business set, and generating the search result list according to the ranking result, comprises: Determining a composite score for each business in the candidate business set based on the score for the first target business, the score for the second target business, and the score for the third target business; and sorting all enterprises in the candidate enterprise set in a descending order according to the comprehensive scores to obtain a retrieval result list.
3. The business name retrieval method according to claim 1, wherein the identifying the second target business containing the character string of the business query text comprises: Intercepting the enterprise names of the rest enterprises except the first target enterprise word by word from the end character to generate a plurality of character strings with increasing lengths until the whole enterprise name is contained; And determining a second target enterprise of which the enterprise name contains the structured semantic data corresponding to the enterprise query text from the rest enterprises except the first target enterprise.
4. The business name retrieval method according to claim 3, wherein said determining that the business name contains the second target business of the structured semantic data corresponding to the business query text comprises: the method comprises the steps of obtaining a search index, wherein the search index comprises that each enterprise name generates all possible continuous character strings from end characters forward word by word Fu Jiequ; and querying that the enterprise containing the structured semantic data corresponding to the enterprise query text is a second target enterprise according to the retrieval index.
5. The business name retrieval method according to claim 1, wherein the querying a third target business for which the business name is semantically matched with the business query text comprises: And determining a third target enterprise with the enterprise name matched with the structured semantic data in the rest enterprises except the first target enterprise and the second target enterprise according to the structured semantic data corresponding to the enterprise query text, wherein the structured semantic data further comprises at least one of a place field, a business direction field and an organization type field.
6. The business name retrieval method according to claim 5, wherein said assigning a score to said third target business comprises: acquiring the matching degree of the enterprise name of the third target enterprise and the structural semantic data; And determining the score of the third target enterprise according to the matching degree and the preset field weight, wherein the weight of the core name field is larger than the weight of the business direction field, and the weight of the business direction field is larger than or equal to the weights of the place field and the organization type field.
7. The method for retrieving an enterprise name according to claim 1, wherein the obtaining the structured semantic data corresponding to the enterprise query text includes: And analyzing the enterprise query text based on the pre-trained enterprise name word segmentation model to obtain the structural semantic data.
8. The business name retrieval method according to claim 1, wherein the business knowledge graph is constructed by: Acquiring a corresponding relation between an enterprise brand name and an enterprise registration name; establishing association relations among the enterprise brand names, the enterprise registration names, the associated parent companies and the subsidiary companies; And taking the enterprise brand name as a label, and associating the label with the corresponding parent company and subsidiary company entity to form the enterprise knowledge graph.
9. An enterprise name retrieval apparatus, the apparatus comprising: the query text receiving module is used for receiving enterprise query text input by a user and obtaining structured semantic data corresponding to the enterprise query text, wherein the structured semantic data comprises a core name field; The enterprise set acquisition module is used for acquiring an enterprise knowledge graph and inquiring a candidate enterprise set associated with enterprise inquiry text in the enterprise knowledge graph based on the core name field; the first target enterprise determining module is used for querying a first target enterprise consistent with the enterprise query text in the candidate enterprise set and giving a score to the first target enterprise; a second target enterprise determining module, configured to identify, among remaining enterprises of the candidate enterprise set that are divided by the first target enterprise, a second target enterprise that includes a character string of the enterprise query text, and assign a score to the second target enterprise; A third target enterprise determining module, configured to query, among remaining enterprises of the candidate enterprise set that are divided by the first target enterprise and the second target enterprise, a third target enterprise whose enterprise name is semantically matched with the enterprise query text, and assign a score to the third target enterprise; and the search result generation module is used for sorting according to the scores of all enterprises in the candidate enterprise set and generating a search result list according to the sorting results.
10. An electronic device comprising a processor, a memory and a computer program stored on the memory and executable on the processor, which when executed by the processor, implements the steps of the business name retrieval method according to any of claims 1-8.
11. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a program which, when executed by a processor, implements the steps of the business name retrieval method according to any of claims 1-8.

Description

Enterprise name retrieval method and device, electronic equipment and storage medium Technical Field The present invention relates to a method for searching a business name, a device for searching a business name, an electronic device, and a computer readable storage medium. Background Matching the candidate's business context by work unit name is a critical screening tool when retrieving enterprise information. The current mainstream technical route generally follows "keyword extraction-inverted index-similarity calculation", namely, a universal word segmentation device (such as an IK word segmentation device) is used for segmenting enterprise names to extract keywords, then the retrieval is carried out based on the inverted index, and the correlation ranking is carried out on the results by adopting a statistical model such as TF-IDF or BM 25. However, the existing method only depends on keyword face matching, lacks deep understanding of the association relationship between the user search intention and enterprises, and cannot identify the subordinate and brand relationship between the enterprises. Users often use brand names, product names or simply replace business registration names, but the word segmentation device of the existing method is insufficient in processing capacity of brand names, product names and other special terms or uncommon words, and is prone to segmentation errors, so that accuracy of search results is low, and user experience is affected. Disclosure of Invention In view of the above, embodiments of the present invention have been made to provide an enterprise name retrieval method, an enterprise name retrieval apparatus, an electronic device, and a computer-readable storage medium that overcome or at least partially solve the above problems. In order to solve the above problem, a first aspect of an embodiment of the present invention provides an enterprise name retrieval method, including: receiving enterprise query text input by a user, and acquiring structural semantic data corresponding to the enterprise query text, wherein the structural semantic data comprises a core name field; acquiring an enterprise knowledge graph, and inquiring a candidate enterprise set associated with the enterprise inquiry text in the enterprise knowledge graph based on the core name field; Querying a first target enterprise consistent with the enterprise query text in the candidate enterprise set, and assigning a score to the first target enterprise; identifying a second target enterprise containing a character string of the enterprise query text in the remaining enterprises of the candidate enterprise set except the first target enterprise, and assigning a score to the second target enterprise; querying a third target enterprise with enterprise names semantically matched with the enterprise query text in the rest enterprises of the candidate enterprise set except the first target enterprise and the second target enterprise, and assigning scores to the third target enterprise; And sorting according to the scores of all enterprises in the candidate enterprise set, and generating a retrieval result list according to the sorting result. Optionally, the ranking according to the scores of the enterprises in the candidate enterprise set, and generating a search result list according to the ranking result, includes: Determining a composite score for each business in the candidate business set based on the score for the first target business, the score for the second target business, and the score for the third target business; and sorting all enterprises in the candidate enterprise set in a descending order according to the comprehensive scores to obtain a retrieval result list. Optionally, the identifying the second target enterprise including the character string of the enterprise query text includes: Intercepting the enterprise names of the rest enterprises except the first target enterprise word by word from the end character to generate a plurality of character strings with increasing lengths until the whole enterprise name is contained; And determining a second target enterprise with enterprise names containing the structured semantic data corresponding to the enterprise query text from the rest enterprises except the first target enterprise. Optionally, the determining that the enterprise name includes the second target enterprise of the structured semantic data corresponding to the enterprise query text includes: the method comprises the steps of obtaining a search index, wherein the search index comprises that each enterprise name generates all possible continuous character strings from end characters forward word by word Fu Jiequ; and querying that the enterprise containing the structured semantic data corresponding to the enterprise query text is a second target enterprise according to the retrieval index. Optionally, the third target enterprise with the query enterprise name semantically matched