A Review on WordNet and Vector Space Analysis for Short-text Semantic Similarity

Abstract

Abstract-Meaningful sentences are the combination of meaning words, if a system wants to process natural language itshould have essential knowledge regarding words and their meanings. The assessment of semantic similarity between the words of a short text is one of the challenging task knowledge based tasks and the tasks in NLP like text summarization, information retrieval, search, categorization of text and machine learning etc. which uses the sentence similarity measures for assessing the similarity between the short-text or sentences. In this paper the survey of two techniques is done which are helpful in generating the extractive text summaries WordNet and Vector space analysis. In vector space model words can be represented as numeric vectors based on different semantic similarity measures, the similarity between the word numeric vectors can be calculated with the semantic measures called WordNet. The information regardin the word and teir meaning in earlier days was provided with the help of traditional dictionaries, but these dictionaries were only helpful for human readers not for machine, WordNet provide combination of traditional lexicographic information and modern information. It is a online lexical database designed for use under the program control, it uses the measure like for calculating the semantic similarity between the concepts. Nouns, verbs, adjectives, adverbs are organized into set of synonyms and semantic relationship between the synonym sets called as Synset.

Short text similarity measurement methods: a review

“…Other studies focus on only several methods of text similarity. Kaundal and Kaur (2017) published a review of measuring short text semantic similarity (STSS) by using two techniques which are vector space model and knowledge- based model by incorporating WordNet. Altszyler et al (2016) focus on the comparison between LSA and word2vec using small corpora.…”