Proceedings of the First International Conference on Combinatorial and Optimization, ICCAP 2021, December 7-8 2021, Chennai, India

Research Article

Semantic Similarity Assessment using Universal Networking Language

Download264 downloads
  • @INPROCEEDINGS{10.4108/eai.7-12-2021.2314540,
        author={A  Chitra and Anupriya  Rajkumar},
        title={Semantic Similarity Assessment using Universal Networking Language},
        proceedings={Proceedings of the First International Conference on Combinatorial and Optimization, ICCAP 2021, December 7-8 2021, Chennai, India},
        publisher={EAI},
        proceedings_a={ICCAP},
        year={2021},
        month={12},
        keywords={paraphrase recognition cross language similarity support vector machine},
        doi={10.4108/eai.7-12-2021.2314540}
    }
    
  • A Chitra
    Anupriya Rajkumar
    Year: 2021
    Semantic Similarity Assessment using Universal Networking Language
    ICCAP
    EAI
    DOI: 10.4108/eai.7-12-2021.2314540
A Chitra1,*, Anupriya Rajkumar1
  • 1: PSG College of Technology
*Contact email: ctr.psg@gmail.com

Abstract

Semantic similarity assessment is a key problem in Natural Language Understanding which finds wide application in Information Retrieval and Extraction. Determining whether two natural language text units are semantically equivalent is a challenging task. In this work, a machine learning approach based on matching of Universal Networking Language (UNL) forms has been proposed for semantic similarity assessment. Features which measure the relatedness of the UNL forms are used as input to a Support Vector Machine classifier to determine semantic equivalence. The performance of the system has been evaluated on the Microsoft Research Paraphrase Corpus with an accuracy of 71%. The suitability of the UNL matching scheme for handling multi-lingual inputs has also been demonstrated.