Dictionary-based sememe knowledge base construction method and device

Abstract

The invention provides a dictionary-based sememe knowledge base construction method and device, and the method comprises the steps: constructing a sememe set according to a controlled word list of a target language dictionary; obtaining a paraphrase word set corresponding to the paraphrase of each semantic item according to the semantic item of each word in the target language dictionary; and according to the sememe set, performing sememe extraction on the paraphrase word set, and according to a sememe extraction result, constructing a sememe knowledge base corresponding to the target language dictionary. Through the dictionary of the target language and the controlled word list corresponding to the dictionary, the primitive knowledge base can be efficiently, economically and automatically constructed for the target language, the problem that time and labor are wasted when the primitive knowledge base is manually constructed is solved, and good practicability is achieved.

Description

technical field [0001] The invention relates to the technical field of natural language processing, in particular to a method and device for constructing a dictionary-based sememe knowledge base. Background technique [0002] In linguistics, a word is defined as the smallest meaningful unit that can be used independently, but not the smallest indivisible semantic unit. Thus, words can be further subdivided into smaller semantic elements. For example, "woman" is further split into "human", "female", and "adult". The smallest indivisible semantic unit in human language is called a sememe. Some linguists believe that the semantics of all words, as well as other concepts, can be represented by a finite set of sememes. Through sememe, words can be analyzed in a more fine-grained manner, which can further help people understand the nature of language. [0003] However, for most human languages, sememes are often very obscure. At present, the sememe knowledge base is mainly con...

AI Technical Summary

Problems solved by technology

[0004] However, at present, most languages ​​do not have a sememe knowledge base similar to HowNet, and building a sememe knowledge base by manual annotation is time-consuming and laborious (for example, HowNet has spent more than 10 years of hard work by several linguistic experts)

Images