MPSC: A Multiple-Perspective Semantics-Crossover Model for Matching Sentences

Abstract

Sentence matching is crucial to many natural language processing (NLP) tasks. Generally, the degree of matching is measured from either of the two perspectives: topic-based match or semantic-based match. The former is to investigate if two sentences discuss the same topic, and the latter performs a deep level semantic matching of texts, which is currently the highlight in research. Deep semantic matching requires adequate modeling from the internal structure of the language objects as well as their interactions. To achieve this goal, this paper proposes a multiple-perspective semantics-crossover (MPSC) model for modeling the semantic-based match of two sentences. The model extracts the matching information of two sentences from the semantic interaction information generated from different angles, so as to calculate the matching degree of the two sentences. The MPSC model not only captures rich matching patterns at different levels but also acquires interactive features from different semantic angles. It can be used to address some important issues in NLP fields, such as information matching in text retrieval, question-answer matching in the Q&A system, and so on. The experimental results show that our proposed model of MPSC has better effectiveness than some popular semantic matching approaches.INDEX TERMS Natural language processing, neural networks, semantics matching, text analysis.

Hierarchical and Multiple-Perspective Interaction Network for Long Text Matching

“…Hu et al [24] proposed a context-aware crossattention mechanism that focuses on contextual information during interaction to achieve better semantic alignment. In contrast, Wang et al [28] did not perform semantic alignment of words or phrases but used full-matching, maxpoolingmatching, attentive-matching, and max-attentive-matching to achieve interaction between texts, and several studies have demonstrated the exceptional performance of this multiperspective matching mechanism [39]. Pre-training based approaches using transformer structure for supervised or unsupervised training on large scale data, represented by BERT and RoBERTa, have achieved very excellent results in the field of text matching.…”

Modelling Simple and Efficient Data Transformation Scheme for Improving Natural Language Processing

“…Park et al [18] have presented a visual analytical system where a bipolar concept has been introduced in modeling. Peng et al [19] have presented a deep semantic approach over multiple perspectives in order to extract the correlation between two sentences. Prakash and Murthy [20] have presented a language specific approach for improving the performance of text-to-speech application.…”