Control method and device of Chinese entity relationship extraction based on word co-occurrence

Abstract

The invention provides a control method of Chinese entity relationship extraction based on word co-occurrence, which comprises the steps of: a. calculating a word correlation degree by statistics of co-occurrence frequency of words; b. calculating word similarity according to the word correlation degree; and c. determining an entity relationship according to the word similarity. The invention also provides a corresponding control device. A corpus used by the method is a news corpus, so that the corpus can be directly created by extracting news texts and titles via the current mature webpage analysis technology without a large amount of manpower participating in the corpus creation; and the method is capable of obtaining such information as word frequency used for calculating the word correlation degree, word position considered when calculating the matching similarity, part of speech of the words and whether the words are verbs, and the like, by utilizing a shallow language rule, for example, participle and part of speech marking in natural language processing and simple statistical techniques, and the method can combine the semantic information of the words with a traditional mode matching method.

Description

technical field [0001] The invention relates to the technical field of entity relationship extraction, in particular to a Chinese entity relationship extraction method based on word co-occurrence and pattern matching. Background technique [0002] The rapid development of the Internet has brought us massive information resources, which are causing profound changes in the way people work, study, live and entertain. However, there are also certain disadvantages. In the face of large-scale information, it is difficult for users to find their real needs. Most of the current search engines are just simple keyword matching, returning many pages with little relevance to users, so users spend a lot of time looking for useful information. Users hope that the search engine can have human-like associative ability, and hope that on the basis of understanding the content of a single concept, it can also find out other information related to this concept, such as the relationship between...

AI Technical Summary

Problems solved by technology

[0003] At present, the methods of entity relationship extraction mainly include: (1) template matching method, which uses natural language processing related knowledge to construct a pattern set and store it. When performing relationship extraction, the preprocessed sentence and pattern set If the matching is successful, it can be considered that the statement has the relational attributes of the corresponding pattern. However, extracting sufficient and high-quality templates is a bottleneck in the application of this method; (2) Based on the dictionary-driven method, the method It is only limited to the case of verb-centered relations; (3) Ontology-based relation extraction, which requires experts to build a large-scale knowledge base, which increases a lot of manual participation; (4) Based on machine learning, this method Training data and using various learning methods to convert into classification problems, usually constructing feature vectors, but this method is relatively slow and not suitable for the extraction of Chinese relations

At the same time, most of the current relationship extraction methods are more suitable for English entity relationship extraction. The Chinese language structure is flexible and the potential meaning is rich. The final recall rate and accuracy rate applied in Chinese entity relationship extraction are not ideal.

Images