Hybrid Deep Learning Model for Sarcasm Detection in Indian Indigenous Language Using Word-Emoji Embeddings

Abstract

Automated sarcasm detection is deemed as a complex natural language processing task and extending it to a morphologically-rich and free-order dominant indigenous Indian language Hindi is another challenge in itself. The scarcity of resources and tools such as annotated corpora, lexicons, dependency parser, Part-of-Speech tagger and benchmark datasets engorge the linguistic challenges of sarcasm detection in low-resource languages like Hindi. Furthermore, as context incongruity is imperative to detect sarcasm, various linguistic, aural and visual cues can be used to predict target utterance as sarcastic. While pre-trained word embeddings capture the meanings, semantic relationships and different types of contexts in the form of word representations, emojis can also render useful contextual information, analogous to human facial expressions, for gauging sarcasm. Thus, the goal of this research is to demonstrate the use of a hybrid deep learning model trained using two embeddings, namely word and emoji embeddings to detect sarcasm. The model is validated on a Hindi tweets dataset, Sarc-H, manually annotated with sarcastic and non-sarcastic labels. The preliminary results clearly depict the importance of using emojis for sarcasm detection, with our model attaining an accuracy of 97.35% with an F-score of 0.9708. The research validates that automated feature engineering facilitates efficient and repeatable predictive model for detecting sarcasm in indigenous, low-resource languages.

A Survey of Sentiment Analysis and Sarcasm Detection

“…The study evaluates the model using real-world datasets from Twitter and Reddit, demonstrating its applicability and potential in practical and business contexts. Kumar et al [26] propose a hybrid deep-learning approach for detecting sarcasm in Hindi text and addressing challenges in languages with limited resources. The model utilized word-emojis embedding and demon-strated high accuracy and F-score in detecting sarcasm in Hindi tweets.…”

“…This suggests that the model has a good capability to understand and classify sarcastic expressions in English tweets, and the model in [22] using CNN-LSTM with BERT for English News Headline sarcasm detection achieved a very high accuracy of 99.63%, showcasing the power of BERT embeddings in sentiment analysis tasks. For notable performance, models like the Hindi Twitter sarcasm detection model in [26] using CNN-LSTM with (CBOW) and Emoji Embedding achieved an accuracy of 97.35%, indicating its effectiveness in handling sarcasm detection in Hindi tweets, and the sarcasm detection model in [25] for English news headlines, using an ensemble model with fuzzy logic and combining Word2Vec, GloVe, and BERT embeddings, achieved an accuracy of 90.81%. For Multilingual Performance: The model in [35] for the Hindi-English code-mixed dataset, using Random Forest (RF) or Logistic Regression (LR) with features like sarcasm tokens, emoticons, and n-grams, achieved an accuracy of 96%, indicating effectiveness in handling multilingual sentiment analysis.…”

“…Sarcasm [14][15][16][17][18], a statement conveying the as it can invert the true sentiment of a statement. Despite being a popular research topic, sarcasm detection [19][20][21][22][23][24][25][26] is crucial as sentiment analysis can misinterpret sarcastic sentences, leading to inaccurate sentiment classifications. The difficulty lies in the nuanced nature of human emotions and expressions conveyed through text, making automatic sarcasm detection a challenging task within natural language processing (NLP) [27,28].…”

Deep Learning for Sarcasm Identification in News Headlines

“…Reference [23] proposed a hybrid deep learning model for sarcasm detection in Indian indigenous languages using word-emoji embeddings. The model is a supervised learning model that uses a combination of word embeddings and emoji embeddings to detect sarcasm.…”

“…Kumar, S.R. Sangwan et al [23] Sarc-H dataset with a total of 1004 tweets was built with 414 tweets labelled as sarcastic and 590 labelled as non-sarcastic CNN-LSTM 91%…”