Short Text Semantic Similarity Measurement Approach Based on Semantic Network

Estimating the semantic similarity between short texts plays an increasingly prominent role in many fields related to text mining and natural language processing applications, especially with large increase volume of textual data that is produced daily. Traditional approaches for calculating degree two texts, based on words they share, do not perform well because similar may be written different terms by employing synonyms. As a result, should semantically compared. In this paper, measurement method presented which combines knowledge-based corpus-based information build network represents relationship compared extracts them. Representing as best knowledge representation comes close human mind's understanding where reflects sentence's semantic, syntactical, structural knowledge. The visual objects, their qualities, relationships. WordNet lexical database has been used source while GloVe pre-trained word embedding vectors have source. proposed was tested using three datasets, DSCS, SICK, MOHLER datasets. A good result obtained RMSE MAE.