|
Vol.10, No.1, February 2021. ISSN: 2217-8309 eISSN: 2217-8333
TEM Journal
TECHNOLOGY, EDUCATION, MANAGEMENT, INFORMATICS Association for Information Communication Technology Education and Science |
ABSA: Computational Measurement Analysis Approach for Prognosticated Aspect Extraction System
Maganti Syamala, N.J.Nalini
© 2021 Maganti Syamala, published by UIKTEN. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. (CC BY-NC-ND 4.0)
Citation Information: TEM Journal. Volume 10, Issue 1, Pages 82-94, ISSN 2217-8309, DOI: 10.18421/TEM101-11, February 2021.
Received: 13 September 2020. Revised: 26 December 2020.
Abstract:
Aspect based sentient analysis (ABSA) is identified as one of the current research problems in Natural Language Processing (NLP). Traditional ABSA requires manual aspect assignment for aspect extraction and sentiment analysis. In this paper, to automate the process, a domain-independent dynamic ABSA model by the fusion of Efficient Named Entity Recognition (E-NER) guided dependency parsing technique with Neural Networks (NN) is proposed. The extracted aspects and sentiment terms by E-NER are trained to a Convolutional Neural Network (CNN) using Word embedding’s technique. Aspect categorybased polarity prediction is evaluated using NLTK Vader Sentiment package. The proposed model was compared to traditional rule-based approach, and the proposed dynamic model proved to yield better results by 17% when validated in terms of correctly classified instances, accuracy, precision, recall and F-Score using machine learning algorithms.
Keywords –Aspect, Category, Extraction, Dependency Parsing, Domain-Independent, Dynamic, Named Entity Recognition, Polarity, Prediction, Sentiment. |
|
----------------------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------------------- |