Chinese entity linking method integrated with pinyin information

A Chinese entity linking method integrated with pinyin information comprises the following steps that (1) a local knowledge base is constructed, the knowledge base should contain a plurality of entities, and each entity corresponds to a unique identification id, an alias and a related descriptive text; 2) selecting training data, and performing corresponding format processing; 3) named entity identification; 4) generating candidate entities; and 5) entity ambiguity elimination: mainly based on a dichotomy thought, fusing pinyin information in the step, then sorting according to probabilities of candidate entities, and taking the highest probability as a correct entity. The method has the beneficial effect that the problem of homomorphic, abnormal-sound and abnormal-sense entity link in the Chinese entity can be well solved. 一种融入拼音信息的中文实体链接方法，包括以下步骤：1)构建本地知识库，知识库中应当包含很多实体，每一个实体对应着唯一的标识id、别名以及相关的描述性文本；2)选取训练数据，并进行相应格式处理；3)命名实体识别；4)候选实体生成；5)实体消岐，主要基于二分类思想，在此步骤将拼音信息融入，然后按照候选实体的概率排序，取最高概率为正确实体。本发明的有益效果为：可以很好的解决中文实