Deep learning and multilingual sentiment analysis on social media data: An overview

Abstract Sentiment analysis is a process in Natural Language Processing that involves detecting and classifying emotions in texts. The emotion is focused on a specific thing, an object, an incident, or an individual. Although some tasks are concerned with detecting the existence of emotion in text, others are concerned with finding the polarities of the text, which is classified as positive, negative, or neutral. The task of determining whether a comment contains inappropriate text that affects either individual or group is called offensive language identification. The existing research has concentrated more on sentiment analysis and offensive language identification in a monolingual data set than code-mixed data. Code-mixed data is framed by combining words and phrases from two or more distinct languages in a single text. It is quite challenging to identify emotion or offensive terms in the comments since noise exists in code-mixed data. The majority of advancements in hostile language detection and sentiment analysis are made on monolingual data for languages with high resource requirements. The proposed system attempts to perform both sentiment analysis and offensive language identification for low resource code-mixed data in Tamil and English using machine learning, deep learning and pre-trained models like BERT, RoBERTa and adapter-BERT. The dataset utilized for this research work is taken from a shared task on Multi task learning DravidianLangTech@ACL2022. Another challenge addressed by this work is the extraction of semantically meaningful information from code-mixed data using word embedding. The result represents an adapter-BERT model gives a better accuracy of 65% for sentiment analysis and 79% for offensive language identification when compared with other trained models. 

/pdf/deep-learning-based-sentiment-analysis-and-offensive-3u3a873x.pdf

Deep learning based sentiment analysis and offensive language identification on multilingual code-mixed data

Emotion detection (ED) in tweets is a text classification problem that is of interest to Natural Language Processing (NLP) researchers. Code-mixing (CM) is a process of mixing linguistic units such as words of two different languages. The CM languages are characteristically different from the languages whose linguistic units are used for mixing. Whilst NLP has been shown to be successful for low-resource languages, it becomes challenging to perform NLP tasks on CM languages. As for ED, it has been rarely investigated on CM languages such as Hindi—English due to the lack of training data that is required for today’s data-driven classification algorithms. This research proposes a gold standard dataset for detecting emotions in CM Hindi–English tweets. This paper also presents our results about the investigation of the usefulness of our gold-standard dataset while testing a number of state-of-the-art classification algorithms. We found that the ED classifier built using SVM provided us the highest accuracy (75.17%) on the hold-out test set. This research would benefit the NLP community in detecting emotions from social media platforms in multilingual societies.

Identifying Emotions in Code Mixed Hindi-English Tweets

As the innovation age encourages individuals to communicate their emotions via web-based media locales like Facebook, Instagram, Twitter, and so on. Numerous individuals share their musings and thoughts step by step in twitter similar to other tweets. This is considered as an exciting and attractive way to express ourselves as inspections are increasing bit by bit, due to which rundown of surveys are necessary for the job where summed up of text is required to give helpful data from the huge number of surveys. It is exceptionally hard for an individual to extricate helpful information or sum up it from the extremely enormous record. This paper focuses on comparison and analysis of different sentiment analysis techniques. It gives a far-reaching review about the recent and past examinations on sentiment analysis, challenges, and approaches for future angles.

Twitter data in Emotional Analysis - A study

Now a day’s most of the political leaders use social media to easily communicate with the people, such as sharing their ideas, promoting their policies etc. Contextual text mining is used to acknowledge the opinions of the political leaders as well as attitude on different subjects such as opinions, discussions and microblogs. Natural language processing (NLP) is included for performing the contextual text mining in order to provide the communication between the human and the machine with natural language. A new model has been implemented to compare the dataset of the political leaders which is extracted from facebook and twitter with different machine learning algorithms like Support Vector Machine (SVM), Naive Bayes Classifier (NBC) and Ensemble Learning Methods (ELM) to provide better and accurate results than other machine learning algorithms.

Contextual Text Mining on Social Media of Political Leaders Using Machine Learning Algorithms

Given the wide applicability of social media platforms, individuals increasingly turn to the web to seek and share information, opinions, comments and suggestions which results in proliferation of user generated large volume of text data available for interpretation. A large number of users in India write their feelings or emotions in more than one language, thereby a large volume of text data is made available for Natural Language Processing (NLP) researchers. Sentiment Analysis (SA) of code-mixed text provides useful information in the field of politics, marketing, business, health, sports and what not. During the past decade the work on Sentiment Analysis of Indian language textual data, particularly in Hindi has got momentum in contrast to code-mixed Indian language text. However, due to non-availability of language and vocabulary (linguistic and lexical) tools and annotated resources, the task of Sentiment Analysis of Indian Languages becomes somehow difficult. In this study an attempt has been made to provide a detailed summary of Sentiment Analysis of Indian languages with a special focus on code mixed Indian Languages.

Review on Sentiment Analysis of Indian Languages with a Special Focus on Code Mixed Indian Languages

Medical images often get affected by noise and artifacts while acquisition techniques or patient’s movement. Some time radiologists unable to make useful and accurate conclusion from noisy images. The main aim of this study is to conduct a comparative measure of various denoising techniques such as Gaussian filter, Median filter, Bilateral filter, Non-Local Mean filter(NLM), Total Variation (TV) Anisotropic Diffusion (AD), BM3D method, on brain NCCT(Non-Contrast computed tomography) images. The prime focus of this research is to make comparative analysis of these techniques not merely to remove but also to preserve edges of brain NCCT images. The experimental results present that BM3D shows the best performance in terms of PSNR value, followed by total variation method and anisotropic diffusion method in terms of removal of noise. However marginal loss of edges and other fine details are there.

Nikita Nikita

Papers

Review on Sentiment Analysis of Indian Languages with a Special Focus on Code Mixed Indian Languages

Comparative study of different denoising techniques on Brain NCCT images