THE SIGNIFICANCE OF REGIONAL LANGUAGE TEXT SENTIMENT ANALYSIS, EMOTION DETECTION, AND NAMED ENTITY RECOGNITION FOR MARATHI LANGUAGE TEXT: IMPROVING UNDERSTANDING AND PRECISION.
Abstract
Performing regional language sentiment analysis, Marathi language emotion detection, and named entity recognition (NER) and emotional topic modeling on Marathi language text is of utmost importance for various reasons. Marathi, being one of the prominent languages spoken in India, encompasses a vast amount of textual data ranging from social media posts to news articles. The analysis of this data provides valuable insights into the collective sentiments, emotions, and significant entities that shape public opinion and discourse. Sentiment Analysis involves determining whether an input text is positive, negative or neutral as in polarity check. In the context of Marathi language text, sentiment analysis can be particularly advantageous for businesses to assess customer satisfaction, for politicians to comprehend public opinion, and for social researchers to study societal trends. The ability to automatically analyze sentiments aids in real-time monitoring and decision-making. Regional Language Emotion Detection analysis provides a general classification of text in seven different emotions defined in the language resources like dictionary, emotion detection delves deeper to identify specific emotions such as joy, anger, sadness, and surprise. For Marathi texts, this can enhance the understanding of how certain events or topics emotionally resonate with the population. For instance, media organizations can utilize emotion detection to tailor their content more effectively to their audience's emotional responses. Named Entity Recognition for Marathi language involves identifying and categorizing tokens as names of person, organizations, cities, animals and other significant terms within a text. In Marathi, NER can assist in structuring unorganized text data, making it easier to extract meaningful information. This is particularly valuable in news aggregation, automated reporting, and database management, where promptly identifying key entities can save considerable time and resources. Ensuring the precision of NLP tasks is of utmost importance. The evaluation of sentiment analysis, emotion detection, and NER models should be conducted rigorously using metrics like precision scores, recall factor and F1-score. To achieve high accuracy in Marathi text, it is crucial to create robust datasets and employ cross-validation techniques. By developing accurate NLP models in Marathi, we can bridge the digital divide and extend the benefits of technology to non-English-speaking populations, promoting inclusivity. To performing sentiment analysis, emotion detection, and NER on Marathi text allows for better extraction, understanding, and utilization of the extensive and valuable textual data available in this language. Accurate NLP models in Marathi not only enhance user engagement and satisfaction but also facilitate informed decision-making across various sectors.
Keyword : Natural Language Processing, Marathi Language Dataset, Regional Language Sentiment Analysis, Regional Language Named Entity Recognition, Regional Language Emotion Detection, Sentiment Analysis, Web Scrapping
This work is licensed under a Creative Commons Attribution 4.0 International License.
References
[1] Ansari, Mohammed Arshad and Govilkar, Sharvari, Sentiment Analysis of Mixed Code for the Transliterated Hindi and Marathi Texts (2018). International Journal on Natural Language Computing (IJNLC) Vol. 7, No.2, April 2018, Available at SSRN: https://ssrn.com/abstract=3429694 or http://dx.doi.org/10.2139/ssrn.3429694 [2] Bolaj, Pooja & Govilkar, Sharvari. (2016). Text Classification for Marathi Documents using Supervised Learning Methods. International Journal of Computer Applications. 155. 6-10. 10.5120/ijca2016912374. [3] Ratna Nitin Patil, Yadvendra Pratap Singh, Shitalkumar Adhar Rawandale, Sofia Singh,Improving Sentiment Classification on Restaurant Reviews Using Deep Learning Models, Procedia Computer Science, Volume 235,2024,Pages 3246-3256,ISSN 1877-0509,https://doi.org/10.1016/j.procs.2024.04.307. [4] Shivani Jain, Anju Saha, Improving and comparing performance of machine learning classifiers optimized by swarm intelligent algorithms for code smell detection, Science of Computer Programming, Volume 237, 2024, 103140, ISSN 0167-6423, https://doi.org/10.1016/j.scico.2024.103140. [5] Abdulfattah Ba Alawi, Ferhat Bozkurt, A hybrid machine learning model for sentiment analysis and satisfaction assessment with Turkish universities using Twitter data, Decision Analytics Journal,Volume 11, 2024, 100473, ISSN 2772-6622, https://doi.org/10.1016/j.dajour.2024.100473. [6] Punithavathi Rasappan, Manoharan Premkumar, Garima Sinha, Kumar Chandrasekaran, Transforming sentiment analysis for e-commerce product reviews: Hybrid deep learning model with an innovative term weighting and feature election, Information Processing & Management, Volume 61, Issue 3, 2024, 103654, ISSN 0306-4573, https://doi.org/10.1016/j.ipm.2024.103654. [7] Aruna Gladys A., Vetriselvi V., Sentiment analysis on a low-resource language dataset using multimodal representation learning and cross-lingual transfer learning, Applied Soft Computing, Volume 157, 2024, 111553, ISSN 1568-4946, https://doi.org/10.1016/j.asoc.2024.111553. [8] Linguistic Data Consortium for Indian Languages (LDC-IL)https://www.ldcil.org/ [9] S. Sirsat and N. Zulpe, "Efficient Dataset Preparation Techniques for Regional/Marathi Language Analysis: Creating Customized Dataset for Regional Language/Marathi Language Text Analysis," 2023 Somaiya International Conference on Technology and Information Management (SICTIM), Mumbai, India, 2023, pp. 90-95, doi: 10.1109/SICTIM56495.2023.10104666. URL: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10104666&isnumber=10104656 [10] S. Sirsat and N. Zulpe, "Recognizing Sentiments In Regional Language Text From Social Media Using Machine Learning Approach: A comparative Study to understand the state of the art tools and associated functionalities available," 2023 International Conference on Advanced Computing Technologies and Applications (ICACTA), Mumbai, India, 2023, pp. 1-8, doi: 10.1109/ICACTA58201.2023.10393104. URL: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10393104&isnumber=10391864 [11] S. Sirsat and N. Zulpe, ”Recognizing Sentiments of Maharashtra State Regional Language Marathi Text from Social Media Using Machine Learning techniques”, Journal of Informatics Education and Research, Vol. 4 No. 2 (2024), ISSN: 1526-4726, pg no: 1872,86; Url: https://www.jier.org/index.php/journal/article/view/1004 [12] S. Sirsat and N. Zulpe, “Recognizing Sentiments In Regional Language Text From Social Media Using Machine Learning Approach : the way ahead through cross domain applications”, COCSIT Conference Proceedings, ISSN – 2322-0015