Laiba Mehnaz, Debanjan Mahata, Rakesh Gosangi, Uma Sushmitha Gunturi, Riya Jain, Gauri Gupta, Amardeep Kumar, Isabelle G. Lee, Anish Acharya, Rajiv Ratn Shah. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. 2021.

Rakesh Gosangi, Ravneet Arora, Mohsen Gheisarieha, Debanjan Mahata, Haimin Zhang. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2021.

Twitter is a social media platform where users express opinions over a variety of issues. Posts offering grievances or complaints can be utilized by private/ public organizations to improve their service and promptly gauge a low-cost assessment. In this paper, we propose an iterative methodology which aims to identify complaint based posts pertaining to the transport domain. We perform comprehensive evaluations along with releasing a novel dataset for the research purposes1.

Visual Speech Recognition (VSR) is the process of recognizing or interpreting speech by watching the lip movements of the speaker. Recent machine learning based approaches model VSR as a classification problem; however, the scarcity of training data leads to error-prone systems with very low accuracies in predicting unseen classes. To solve this problem, we present a novel approach to zero-shot learning by generating new classes using Generative Adversarial Networks (GANs), and show how the addition of unseen class samples increases the accuracy of a VSR system by a significant margin of 27% and allows it to handle speaker-independent out-of-vocabulary phrases. We also show that our models are language agnostic and therefore capable of seamlessly generating, using English training data, videos for a new language (Hindi). To the best of our knowledge, this is the first work to show empirical evidence of the use of GANs for generating training samples of unseen classes in the domain of VSR, hence facilitating zero-shot learning. We make the added videos for new classes publicly available along with our code1.

In this paper, we present a keyphrase generation approach using conditional Generative Adversarial Networks (GAN). In our GAN model, the generator outputs a sequence of keyphrases based on the title and abstract of a scientific article. The discriminator learns to distinguish between machine-generated and human-curated keyphrases. We evaluate this approach on standard benchmark datasets. Our model achieves state-of-the-art performance in generation of abstractive keyphrases and is also comparable to the best performing extractive techniques. We also demonstrate that our method generates more diverse keyphrases and make our implementation publicly available1.

We introduce a new keyphrase generation approach using Generative Adversarial Networks (GANs). For a given document, the generator produces a sequence of keyphrases, and the discriminator distinguishes between human-curated and machine-generated keyphrases. We evaluated this approach on standard benchmark datasets. We observed that our model achieves state-of-the-art performance in the generation of abstractive keyphrases and is comparable to the best performing extractive techniques. Although we achieve promising results using GANs, they are not significantly better than the state-of-the-art generative models. To our knowledge, this is one of the first works that use GANs for keyphrase generation. We present a detailed analysis of our observations and expect that these findings would help other researchers to further study the use of GANs for the task of keyphrase generation.

In this paper, we present a new corpus consisting of sentences from Hindi short stories annotated for five different discourse modes argumentative, narrative, descriptive, dialogic and informative. We present a detailed account of the entire data collection and annotation processes. The annotations have a very high inter-annotator agreement (0.87 k-alpha). We analyze the data in terms of label distributions, part of speech tags, and sentence lengths. We characterize the performance of various classification algorithms on this dataset and perform ablation studies to understand the nature of the linguistic models suitable for capturing the nuances of the embedded discourse structures in the presented corpus.

In the era of MOOCs, online exams are taken by millions of candidates, where scoring short answers is an integral part. It becomes intractable to evaluate them by human graders. Thus, a generic automated system capable of grading these responses should be designed and deployed. In this paper, we present a fast, scalable, and accurate approach towards automated Short Answer Scoring (SAS). We propose and explain the design and development of a system for SAS, namely AutoSAS. Given a question along with its graded samples, AutoSAS can learn to grade that prompt successfully. This paper further lays down the features such as lexical diversity, Word2Vec, prompt, and content overlap that plays a pivotal role in building our proposed model. We also present a methodology for indicating the factors responsible for scoring an answer. The trained model is evaluated on an extensively used public dataset, namely Automated Student Assessment Prize Short Answer Scoring (ASAP-SAS). AutoSAS shows state-of-the-art performance and achieves better results by over 8% in some of the question prompts as measured by Quadratic Weighted Kappa (QWK), showing performance comparable to humans.

The availability of large-scale online social data, coupled with computational methods, can help us answer fundamental questions relating to our social lives, particularly our health and well-being.The #MeToo trend has led to people talking about personal experiences of harassment more openly.This work attempts to aggregate such experiences of sexual abuse to facilitate a better understanding of social media constructs and to bring about social change.It has been found that disclosure of abuse has positive psychological impacts.Hence, we contend that such information can be leveraged to create better campaigns for social change by analyzing how users react to these stories and can be used to obtain a better insight into the consequences of sexual abuse.We use a three-part Twitter-Specific Social Media Language Model to segregate personal recollections of sexual harassment from Twitter posts.An extensive comparison with state-of-the-art generic and specific models along with a detailed error analysis explores the merit of our proposed model.* denotes equal contribution.1 https://www.who.int/news-room/fact-sheets/detail/violence-against-women

Technological advancements have led to the creation of social media platforms like Twitter, where people have started voicing their views over rarely discussed and socially stigmatizing issues. Twitter, is increasingly being used for studying psycho-linguistic phenomenon spanning from expressions of adverse drug reactions, depressions, to suicidality. In this work we focus on identifying suicidal posts from Twitter. Towards this objective we take a multipronged approach and implement different neural network models such assequential models andgraph convolutional networks, that are trained on textual content shared in Twitter, the historical tweeting activity of the users and social network formed between different users posting about suicidality. We train a stacked ensemble of classifiers representing different aspects of suicidal tweeting activity, and achieve state-of-the-art results on a new manually annotated dataset developed by us, that contains textual as well as network information of suicidal tweets. We further investigate into the trained models and perform qualitative analysis showing how historical tweeting activity and rich information embedded in the homophily networks amongst users in Twitter, aids in accurately identifying tweets expressing suicidal intent.

Objective: We executed the Social Media Mining for Health (SMM4H) 2017 shared tasks to enable the community-driven development and large-scale evaluation of automatic text processing methods for the classification and normalization of health-related text from social media. An additional objective was to publicly release manually annotated data. Materials and Methods: We organized 3 independent subtasks: automatic classification of self-reports of 1) adverse drug reactions (ADRs) and 2) medication consumption, from medication-mentioning tweets, and 3) normalization of ADR expressions. Training data consisted of 15 717 annotated tweets for (1), 10 260 for (2), and 6650 ADR phrases and identifiers for (3); and exhibited typical properties of social-media-based health-related texts. Systems were evaluated using 9961, 7513, and 2500 instances for the 3 subtasks, respectively. We evaluated performances of classes of methods and ensembles of system combinations following the shared tasks. Results: Among 55 system runs, the best system scores for the 3 subtasks were 0.435 (ADR class F1-score) for subtask-1, 0.693 (micro-averaged F1-score over two classes) for subtask-2, and 88.5% (accuracy) for subtask-3. Ensembles of system combinations obtained best scores of 0.476, 0.702, and 88.7%, outperforming individual systems. Discussion: Among individual systems, support vector machines and convolutional neural networks showed high performance. Performance gains achieved by ensembles of system combinations suggest that such strategies may be suitable for operational systems relying on difficult text classification tasks (eg, subtask-1). Conclusions: Data imbalance and lack of context remain challenges for natural language processing of social media text. Annotated data from the shared task have been made available as reference standards for future studies (http://dx.doi.org/10.17632/rxwfb3tysd.1).

Social media is inarguably a powerful medium for mobilizing support for various real-life events be it for social, political, or economic transformation. Further, in contrast to the generic information obtained from the mainstream media, novel and specific information available at social media sites makes them valuable sources for event analysis. However, due to the power law distribution of the Internet, these overwhelmingly large number of sources are buried in the Long Tail making it extremely challenging to identify the quality sources among them. In this research, we propose an evolutionary mutual reinforcement model to confront these challenges. Due to absence of ground truth, a novel evaluation strategy is introduced. The results indicate tremendous potential. 25% to 130% information gain is obtained with the proposed approach when compared against the state-of-the-art baselines, viz. Google blog search and Icerocket blog search. Further, our ranking methodology is capable of identifying the highly informative sources much earlier than the aforementioned baselines. The proposed model affords an apparatus for micro and macro event analysis.