Open domain news text relationship extraction based on dependency syntax

Entity relation extraction is an important sub-task of information extraction. Its goal is to determine whether there is a semantic relationship between entities. For Chinese, semantic relations are more complex and the effect of entity relation extraction is more significant. Besides, the use of verbs as a relational expression in Chinese is likely to cause ambiguity in the relationship between entities due to the intricate nature of Chinese grammar. Starting from the relationship expression, we propose and implement an open-field Chinese relationship extraction method based on dependency analysis. Dependency analysis is performed on the input single sentence and whether the single sentence is a verb-predicate statement is determined by the dependency arc of the dependency analysis output. If it is a verb-predicate statement, the Chinese grammatical heuristic rule is used to extract the relationship expression. The triad is evaluated according to the position of the argument elements. At last, the qualified triple is output. The experiments on SogouCA and SogouCS corpus show that the proposed method is suitable for large-scale corpus and has good performance and portability. And experimental results F-score is up to 81.6%.