Aditya Joshi, Vinita Sharma, Pushpak Bhattacharyya. Proceedings of the 53rd Annual Meeting Association for Computational Linguistics and 7th International Joint Conference on Natural Language Processing (Volume 2: Short Papers). 2015.

Md Shad Akhtar, Dushyant Chauhan, Deepanway Ghosal, Soujanya Poria, Asif Ekbal, Pushpak Bhattacharyya. Proceedings of the 2019 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019.

This paper makes a simple increment to state-of-the-art in sarcasm detection research. Existing approaches are unable capture subtle forms of context incongruity which lies at the heart sarcasm. We explore if prior work can be enhanced using semantic similarity/discordance between word embeddings. augment embedding-based features four feature sets reported past. also experiment with types observe an improvement detection, irrespective embedding used or original set our augmented. For...

Multi-modal sentiment analysis offers various challenges, one being the effective combination of different input modalities, namely text, visual and acoustic. In this paper, we propose a recurrent neural network based multi-modal attention framework that leverages contextual information for utterance-level prediction. The proposed approach applies on multi-utterance representations tries to learn contributing features amongst them. We evaluate our two benchmark datasets, viz. CMU...

We present the IIT Bombay English-Hindi Parallel Corpus. The corpus is a compilation of parallel corpora previously available in public domain as well new we collected. contains 1.49 million segments, which 694k segments were not domain. has been pre-processed for machine translation, and report baseline phrase-based SMT NMT translation results on this corpus. This used two editions shared tasks at Workshop Asian Language Translation (2016 2017). freely non-commercial research. To best our...

Automatic sarcasm detection is the task of predicting in text. This a crucial step to sentiment analysis, considering prevalence and challenges sentiment-bearing Beginning with an approach that used speech-based features, has witnessed great interest from analysis community. paper first known compilation past work automatic detection. We observe three milestones research so far: semi-supervised pattern extraction identify implicit sentiment, use hashtag-based supervision, context beyond...

In this paper, we hypothesize that sarcasm is closely related to sentiment and emotion, thereby propose a multi-task deep learning framework solve all these three problems simultaneously in multi-modal conversational scenario. We, at first, manually annotate the recently released MUStARD dataset with emotion classes, both implicit explicit. For multi-tasking, two attention mechanisms, viz. Inter-segment Inter-modal Attention (Ie-Attention) Intra-segment (Ia-Attention). The main motivation of...

Cognitive NLP systems- i.e., systems that make use of behavioral data - augment traditional text-based features with cognitive extracted from eye-movement patterns, EEG signals, brain-imaging etc. Such extraction is typically manual. We contend manual may not be the best way to tackle text subtleties characteristically prevail in complex classification tasks like Sentiment Analysis and Sarcasm Detection, even choice should delegated learning system. introduce a framework automatically...

In this paper, we propose a novel method for combining deep learning and classical feature based models using Multi-Layer Perceptron (MLP) network financial sentiment analysis. We develop various on Convolutional Neural Network (CNN), Long Short Term Memory (LSTM) Gated Recurrent Unit (GRU). These are trained top of pre-trained, autoencoder-based, word embeddings lexicon features. An ensemble is constructed by these supervised model Support Vector Regression (SVR). evaluate our proposed...

We propose a multi-task ensemble framework that jointly learns multiple related problems. The model aims to leverage the learned representations of three deep learning models (i.e., CNN, LSTM and GRU) hand-crafted feature representation for predictions. Through framework, we address four problems emotion sentiment analysis, i.e., "emotion <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">classification</i> &...

Sarcasm understanding may require information beyond the text itself, as in case of 'I absolutely love this restaurant!' which be sarcastic, depending on contextual situation.We present first quantitative evidence to show that historical tweets by an author can provide additional context for sarcasm detection.Our detection approach uses two components: a contrast-based predictor (that identifies if there is sentiment contrast within target tweet), and tweet-based expressed towards entity...

Detecting cyberbullying from memes is highly challenging, because of the presence implicit affective content which also often sarcastic, and multi-modality (image + text). The current work first attempt, to best our knowledge, in investigating role sentiment, emotion sarcasm identifying multi-modal a code-mixed language setting. As contribution, we have created benchmark meme dataset called MultiBully annotated with bully, labels collected open-source Twitter Reddit platforms. Moreover,...

With the advent of Internet, large amount digital text is generated everyday in form news articles, research publications, blogs, question answering forums and social media. It important to develop techniques for extracting information automatically from these documents, as lot hidden within them. This extracted can be used improve access management knowledge corpora. Several applications such Question Answering, Information Retrieval would benefit this information. Entities like persons...

This paper is a novel study that views sarcasm detection in dialogue as sequence labeling task, where made up of utterances.We create manuallylabeled dataset from TV series 'Friends' annotated with sarcasm.Our goal to predict each utterance, using sequential nature scene.We show performance gain compared classification-based approaches.Our experiments are based on three sets features, one derived information our dataset, the other two past works.Two algorithms (SVM-HMM and SEARN) outperform...

Dushyant Singh Chauhan, Md Shad Akhtar, Asif Ekbal, Pushpak Bhattacharyya. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint (EMNLP-IJCNLP). 2019.