Entity mining system and method

Abstract

The present invention provides a system and method for entity mining. The system includes: a labeling unit configured to obtain the first group of knowledge information and several entities in the current field, and mark the first group of knowledge information according to the several entities; a rule mining unit , configured to screen several associated high-frequency words from the first set of knowledge information of labeled entities, and to screen some phrases with the highest frequency of occurrence in several high-frequency words to generate orderly association rules; the entity mining unit is configured to obtain For the second group of knowledge information in the current field, mine the second group of knowledge information according to the ordered association rules to obtain several entity candidates; the filtering unit is configured to filter each entity candidate according to a predetermined filtering rule to generate an entity mining result and output. The invention utilizes ordered association rules to carry out entity mining, improves the accuracy rate of entity mining, effectively filters the mining results, and ensures the accuracy of the mining results.

Description

Entity mining system and method technical field [0001] This application relates to the technical field of data mining, in particular to a system and method for entity mining. Background technique At present, in some fields, the entity data stored in the system is not comprehensive enough and perfect, and it is difficult to satisfy user search. The demand for querying information causes users to spend a lot of time searching, reading, and judging and analyzing information. [0003] For example, in the medical field, a patient is learning that he has a certain disease or speculating that he may have a certain disease. When the disease is diagnosed, the corresponding disease information is usually found by means of search, and when the name of the disease is not stored in the system, or only When the scientific name of the disease is stored, but its alias or abbreviation is not stored, the search results will be difficult to meet the user's needs. Therefore, users need ...

AI Technical Summary

Problems solved by technology

[0002] At present, in some fields, the entity data stored in the system is not comprehensive and perfect, and it is difficult to meet the needs of users to search and query information, resulting in users needing to spend a lot of time searching, reading and judging and analyzing information

[0003] For example, in the medical field, when patients know that they have a certain disease or speculate that they may have a certain disease, they usually use search to find the corresponding disease information, but when the name of the disease is not stored in the system, or When only the scientific name of the disease is stored without storing its alias or abbreviation, the search results will be difficult to meet the needs of users, and users need to spend a lot of time and energy searching and judging when searching

[0004] Existing entity mining solutions usually mine entity tags and attributes in structured data pages, and obtain entities through predetermined disordered rules. The accuracy of mining needs to be improved

Images