NITS-Hinglish-SentiMix at SemEval-2020 Task 9: Sentiment Analysis for Code-Mixed Social Media Text Using an Ensemble Model

Proceedings of the 14th International Workshop on Semantic Evaluation, pages 1298–1303 Barcelona, Spain (Online), December 12, 2020. 1298 NITS-Hinglish-SentiMix at SemEval-2020 Task 9: Sentiment Analysis For Code-Mixed Social Media Text Using an Ensemble Model Subhra Jyoti Baroi Nivedita Singh Ringki Das Thoudam Doren Singh Department of Computer Science and Engineering National Institute of Technology Silchar Silchar, Assam -788010 {tushar.barroi99,s.nivedita279,ringkidas,thoudam.doren}@gmail.com Abstract Sentiment Analysis is the process of deciphering what a sentence emotes and classifying them as either positive, negative, or neutral. In recent times, India has seen a huge influx in the number of active social media users and this has led to a plethora of unstructured text data. Since the Indian population is generally fluent in both Hindi and English, they end up generating code-mixed Hinglish social media text i.e. the expressions of Hindi language, written in the Roman script alongside other English words. The ability to adequately comprehend the notions in these texts is truly necessary. Our team, rns2020 participated in Task 9 at SemEval2020 intending to design a system to carry out the sentiment analysis of code-mixed social media text. This work proposes a system named NITS-Hinglish-SentiMix to viably complete the sentiment analysis of such code-mixed Hinglish text. The proposed framework has recorded an F-Score of 0.617 on the test data. 1 Introduction With the advent of social media, India alone stood at 326.1 million users in 2018. A large portion of the population is fluent in both English and Hindi henceforth it is regular to see the use of code-mixed Hinglish language on social media platforms. In the world of natural language processing(NLP), sentiment analysis of the code-mixed Hinglish text is a great challenge. The task at hand is the SemEval-2020 Task 9 (Patwa et al., 2020): Sentiment Analysis for Code-Mixed Social Media Text for Hinglish. Code-mixed Hinglish tweets are provided as training, validation, and test sets for model training, validation, and testing respectively. The proposed framework NITS-Hinglish-SentiMix, an ensemble model wherein different models have been consolidated to improve the general F-Score of the framework. The referenced group model is a mix of a basic LSTM (Long Short Term Memory), an LSTM+Convolution, a BiLSTM (Bidirectional LSTM), and a CNN (Convolution Neural Network) model. This paper is composed as follows, Section 2 is a concise writing overview of related works. Section 3 talks about the strategy which covers bits of insights into the dataset given, alongside the pre-processing, the details on the model architecture. Section 4 talks about the training and results acquired and the paper concludes with Section 5. 2 Literature Survey In recent years, a lot has changed in the field of sentiment analysis. This section discusses a few of the works which helped us build insight into the problem and come up with a new solution. The initial approach towards sentiment analysis includes systems powered by traditional machine learning techniques like the work by (Agarwal and Bhattacharyya, 2005) explores the performance of SVM for the task of multinomial classification of documents. The approach used the …

Content truncated to 500 words