Multi-clue Chinese semantic role labeling based on conditional random fields

With the industrial development of artificial intelligence and Chinese information processing technology, researches on natural language processing is gradually deepening to the level of semantic understanding, and Chinese Semantic Role Labeling is the core technology in the field of semantic understanding. Aiming at the fact that the existing linear labeling model cannot meet the needs of the industrialization of semantic information technology for labeling accuracy, a multi-level linguistic clue combination model optimization method is proposed. First, a benchmark model of conditional random fields with excellent comprehensive labeling performance is selected to build a matching semantic role labeling corpus. Then, multi-level linguistic clues such as morphology and sentence structure are incorporated into the model to achieve multi-level optimization of the model. Finally, by comparing the various indexing experiments, it is demonstrated that the relevant linguistic cues incorporated can effectively enhance the labeling performance of the linear sequence model..