Sentiment Analysis of Mixed Code for the Transliterated Hindi and Marathi Texts

Abstract

The evolution of information Technology has led to the collection of large amount of data, the volume of which has increased to the extent that in last two years the data produced is greater than all the data ever recorded in human history. This has necessitated use of machines to understand, interpret and apply data, without manual involvement. A lot of these texts are available in transliterated code-mixed form, which due to the complexity are very difficult to analyze. The work already performed in this area is progressing at great pace and this work hopes to be a way to push that work further. The designed system is an effort which classifies Hindi as well as Marathi text transliterated (Romanized) documents automatically using supervised learning methods (KNN), Naïve Bayes and Support Vector Machine (SVM)) and ontology based classification; and results are compared to in order to decide which methodology is better suited in handling of these documents. As we will see, the plain machine learning algorithm applications are just as or in many cases are much better in performance than the more analytical approach.

Opinion-Mining on Marglish and Devanagari Comments of YouTube Cookery Channels Using Parametric and Non-Parametric Learning Models

“…Ansari et al [9] introduced an architecture for two code-mix languages, Hindi and Marathi. The architecture included language identification, feature generation and sentiment classification as major steps.…”

Exploring Sentiment Analysis in Indian Regional Languages: Methods, Challenges, and Future Directions

“…These studies have addressed challenges such as informal language, slang, and emoticon usage typical of social media text, with BERT models demonstrating superior performance due to their ability to capture contextual information effectively. The exploration of these models across various languages and contexts emphasizes the dynamic nature of sentiment analysis research and its potential for future advancements [ 1,3,4,5,7,9,10,11,13,14,15,20,22,23,24,30].For languages with limited computational resources, such as Marathi and Urdu, lexicon-based approaches have been proposed as effective methods for sentiment analysis. Researchers have developed lexicons that include lists of positive and negative words, assigning polarity values to facilitate the classification of sentences into sentiments.…”

“…Future work is suggested to focus on expanding the classification capabilities to include figurative language, enriching datasets with more diverse samples, exploring additional algorithms for enhanced accuracy, and further developing sentiment analysis models to accommodate low-resource languages. These directions underscore the evolving nature of sentiment analysis research and its critical role in understanding and leveraging user-generated content in multilingual societies [1,2,3,4,5,7,9,10,11,13,14,15,20,22,23,24,30,31,32].…”

Sentiment analysis on Hindi tweets during COVID‐19 pandemic

“…Translation of words from English to Hindi using search and replace algorithms by Yadav et al 26 have been proposed for similar tasks. Mohammed Ansari et al 27 transliterated to Devanagari by identifying the language and are classified based on POS Tagging. Deep Learning techniques and pretrained word embeddings have widely improved results in the domain of Natural Language Processing.…”