Binny Mathew

Punyajoy Saha, Kiran Garimella, Narla Komal Kalyan et al.

Recently, social media platforms are heavily moderated to prevent the spread of online hate speech, which is usually fertile in toxic words and is directed toward an individual or a community. Owing to such heavy moderation, newer and more subtle techniques are being deployed. One of the most striking among these is fear speech. Fear speech, as the name suggests, attempts to incite fear about a target community. Although subtle, it might be highly effective, often pushing communities toward a physical conflict. Therefore, understanding their prevalence in social media is of paramount importance. This article presents a large-scale study to understand the prevalence of 400K fear speech and over 700K hate speech posts collected from Gab.com. Remarkably, users posting a large number of fear speech accrue more followers and occupy more central positions in social networks than users posting a large number of hate speech. They can also reach out to benign users more effectively than hate speech users through replies, reposts, and mentions. This connects to the fact that, unlike hate speech, fear speech has almost zero toxic content, making it look plausible. Moreover, while fear speech topics mostly portray a community as a perpetrator using a (fake) chain of argumentation, hate speech topics hurl direct multitarget insults, thus pointing to why general users could be more gullible to fear speech. Our findings transcend even to other platforms (Twitter and Facebook) and thus necessitate using sophisticated moderation policies and mass awareness to combat fear speech.

Gaurav Verma, Rynaa Grover, Jiawei Zhou et al. · gatech

Violence-provoking speech -- speech that implicitly or explicitly promotes violence against the members of the targeted community, contributed to a massive surge in anti-Asian crimes during the pandemic. While previous works have characterized and built tools for detecting other forms of harmful speech, like fear speech and hate speech, our work takes a community-centric approach to studying anti-Asian violence-provoking speech. Using data from ~420k Twitter posts spanning a 3-year duration (January 1, 2020 to February 1, 2023), we develop a codebook to characterize anti-Asian violence-provoking speech and collect a community-crowdsourced dataset to facilitate its large-scale detection using state-of-the-art classifiers. We contrast the capabilities of natural language processing classifiers, ranging from BERT-based to LLM-based classifiers, in detecting violence-provoking speech with their capabilities to detect anti-Asian hateful speech. In contrast to prior work that has demonstrated the effectiveness of such classifiers in detecting hateful speech ($F_1 = 0.89$), our work shows that accurate and reliable detection of violence-provoking speech is a challenging task ($F_1 = 0.69$). We discuss the implications of our findings, particularly the need for proactive interventions to support Asian communities during public health crises. The resources related to the study are available at https://claws-lab.github.io/violence-provoking-speech/.

Punyajoy Saha, Kanishk Singh, Adarsh Kumar et al.

Recently, many studies have tried to create generation models to assist counter speakers by providing counterspeech suggestions for combating the explosive proliferation of online hate. However, since these suggestions are from a vanilla generation model, they might not include the appropriate properties required to counter a particular hate speech instance. In this paper, we propose CounterGeDi - an ensemble of generative discriminators (GeDi) to guide the generation of a DialoGPT model toward more polite, detoxified, and emotionally laden counterspeech. We generate counterspeech using three datasets and observe significant improvement across different attribute scores. The politeness and detoxification scores increased by around 15% and 6% respectively, while the emotion in the counterspeech increased by at least 10% across all the datasets. We also experiment with triple-attribute control and observe significant improvement over single attribute results when combining complementing attributes, e.g., politeness, joyfulness and detoxification. In all these experiments, the relevancy of the generated text does not deteriorate due to the application of these controls

Mithun Das, Punyajoy Saha, Binny Mathew et al.

Due to the sheer volume of online hate, the AI and NLP communities have started building models to detect such hateful content. Recently, multilingual hate is a major emerging challenge for automated detection where code-mixing or more than one language have been used for conversation in social media. Typically, hate speech detection models are evaluated by measuring their performance on the held-out test data using metrics such as accuracy and F1-score. While these metrics are useful, it becomes difficult to identify using them where the model is failing, and how to resolve it. To enable more targeted diagnostic insights of such multilingual hate speech models, we introduce a set of functionalities for the purpose of evaluation. We have been inspired to design this kind of functionalities based on real-world conversation on social media. Considering Hindi as a base language, we craft test cases for each functionality. We name our evaluation dataset HateCheckHIn. To illustrate the utility of these functionalities , we test state-of-the-art transformer based m-BERT model and the Perspective API.

Piush Aggarwal, Pranit Chawla, Mithun Das et al.

Exploiting social media to spread hate has tremendously increased over the years. Lately, multi-modal hateful content such as memes has drawn relatively more traction than uni-modal content. Moreover, the availability of implicit content payloads makes them fairly challenging to be detected by existing hateful meme detection systems. In this paper, we present a use case study to analyze such systems' vulnerabilities against external adversarial attacks. We find that even very simple perturbations in uni-modal and multi-modal settings performed by humans with little knowledge about the model can make the existing detection models highly vulnerable. Empirically, we find a noticeable performance drop of as high as 10% in the macro-F1 score for certain attacks. As a remedy, we attempt to boost the model's robustness using contrastive learning as well as an adversarial training-based method - VILLA. Using an ensemble of the above two approaches, in two of our high resolution datasets, we are able to (re)gain back the performance to a large extent for certain attacks. We believe that ours is a first step toward addressing this crucial problem in an adversarial setting and would inspire more such investigations in the future.

Punyajoy Saha, Divyanshu Sheth, Kushal Kedia et al.

Abusive language is a concerning problem in online social media. Past research on detecting abusive language covers different platforms, languages, demographies, etc. However, models trained using these datasets do not perform well in cross-domain evaluation settings. To overcome this, a common strategy is to use a few samples from the target domain to train models to get better performance in that domain (cross-domain few-shot training). However, this might cause the models to overfit the artefacts of those samples. A compelling solution could be to guide the models toward rationales, i.e., spans of text that justify the text's label. This method has been found to improve model performance in the in-domain setting across various NLP tasks. In this paper, we propose RGFS (Rationale-Guided Few-Shot Classification) for abusive language detection. We first build a multitask learning setup to jointly learn rationales, targets, and labels, and find a significant improvement of 6% macro F1 on the rationale detection task over training solely rationale classifiers. We introduce two rationale-integrated BERT-based architectures (the RGFS models) and evaluate our systems over five different abusive language datasets, finding that in the few-shot classification setting, RGFS-based models outperform baseline models by about 7% in macro F1 scores and perform competitively to models finetuned on other source domains. Furthermore, RGFS-based models outperform LIME/SHAP-based approaches in terms of plausibility and are close in performance in terms of faithfulness.

Binny Mathew, Punyajoy Saha, Seid Muhie Yimam et al.

Hate speech is a challenging issue plaguing the online social media. While better models for hate speech detection are continuously being developed, there is little research on the bias and interpretability aspects of hate speech. In this paper, we introduce HateXplain, the first benchmark hate speech dataset covering multiple aspects of the issue. Each post in our dataset is annotated from three different perspectives: the basic, commonly used 3-class classification (i.e., hate, offensive or normal), the target community (i.e., the community that has been the victim of hate speech/offensive speech in the post), and the rationales, i.e., the portions of the post on which their labelling decision (as hate, offensive or normal) is based. We utilize existing state-of-the-art models and observe that even models that perform very well in classification do not score high on explainability metrics like model plausibility and faithfulness. We also observe that models, which utilize the human rationales for training, perform better in reducing unintended bias towards target communities. We have made our code and dataset public at https://github.com/punyajoy/HateXplain

Sai Saketh Aluru, Binny Mathew, Punyajoy Saha et al.

Hate speech detection is a challenging problem with most of the datasets available in only one language: English. In this paper, we conduct a large scale analysis of multilingual hate speech in 9 languages from 16 different sources. We observe that in low resource setting, simple models such as LASER embedding with logistic regression performs the best, while in high resource setting BERT based models perform better. In case of zero-shot classification, languages such as Italian and Portuguese achieve good results. Our proposed framework could be used as an efficient solution for low-resource languages. These models could also act as good baselines for future multilingual hate speech detection tasks. We have made our code and experimental settings public for other researchers at https://github.com/punyajoy/DE-LIMIT.

Punyajoy Saha, Binny Mathew, Pawan Goyal et al.

Reducing hateful and offensive content in online social media pose a dual problem for the moderators. On the one hand, rigid censorship on social media cannot be imposed. On the other, the free flow of such content cannot be allowed. Hence, we require efficient abusive language detection system to detect such harmful content in social media. In this paper, we present our machine learning model, HateMonitor, developed for Hate Speech and Offensive Content Identification in Indo-European Languages (HASOC), a shared task at FIRE 2019. We have used a Gradient Boosting model, along with BERT and LASER embeddings, to make the system language agnostic. Our model came at First position for the German sub-task A. We have also made our model public at https://github.com/punyajoy/HateMonitors-HASOC .

Punyajoy Saha, Binny Mathew, Pawan Goyal et al.

With the online proliferation of hate speech, there is an urgent need for systems that can detect such harmful content. In this paper, We present the machine learning models developed for the Automatic Misogyny Identification (AMI) shared task at EVALITA 2018. We generate three types of features: Sentence Embeddings, TF-IDF Vectors, and BOW Vectors to represent each tweet. These features are then concatenated and fed into the machine learning models. Our model came First for the English Subtask A and Fifth for the English Subtask B. We release our winning model for public use and it's available at https://github.com/punyajoy/Hateminers-EVALITA.

Somnath Banerjee, Maulindu Sarkar, Punyajoy Saha et al.

Recently, influence functions present an apparatus for achieving explainability for deep neural models by quantifying the perturbation of individual train instances that might impact a test prediction. Our objectives in this paper are twofold. First we incorporate influence functions as a feedback into the model to improve its performance. Second, in a dataset extension exercise, using influence functions to automatically identify data points that have been initially `silver' annotated by some existing method and need to be cross-checked (and corrected) by annotators to improve the model performance. To meet these objectives, in this paper, we introduce InfFeed, which uses influence functions to compute the influential instances for a target instance. Toward the first objective, we adjust the label of the target instance based on its influencer(s) label. In doing this, InfFeed outperforms the state-of-the-art baselines (including LLMs) by a maximum macro F1-score margin of almost 4% for hate speech classification, 3.5% for stance classification, and 3% for irony and 2% for sarcasm detection. Toward the second objective we show that manually re-annotating only those silver annotated data points in the extension set that have a negative influence can immensely improve the model performance bringing it very close to the scenario where all the data points in the extension set have gold labels. This allows for huge reduction of the number of data points that need to be manually annotated since out of the silver annotated extension dataset, the influence function scheme picks up ~1/1000 points that need manual correction.

Mithun Das, Rohit Raj, Punyajoy Saha et al.

Hate speech has become one of the most significant issues in modern society, having implications in both the online and the offline world. Due to this, hate speech research has recently gained a lot of traction. However, most of the work has primarily focused on text media with relatively little work on images and even lesser on videos. Thus, early stage automated video moderation techniques are needed to handle the videos that are being uploaded to keep the platform safe and healthy. With a view to detect and remove hateful content from the video sharing platforms, our work focuses on hate video detection using multi-modalities. To this end, we curate ~43 hours of videos from BitChute and manually annotate them as hate or non-hate, along with the frame spans which could explain the labelling decision. To collect the relevant videos we harnessed search keywords from hate lexicons. We observe various cues in images and audio of hateful videos. Further, we build deep learning multi-modal models to classify the hate videos and observe that using all the modalities of the videos improves the overall hate speech detection performance (accuracy=0.798, macro F1-score=0.790) by ~5.7% compared to the best uni-modal model in terms of macro F1 score. In summary, our work takes the first step toward understanding and modeling hateful videos on video hosting platforms such as BitChute.

Mithun Das, Punyajoy Saha, Ritam Dutt et al.

Hate speech is regarded as one of the crucial issues plaguing the online social media. The current literature on hate speech detection leverages primarily the textual content to find hateful posts and subsequently identify hateful users. However, this methodology disregards the social connections between users. In this paper, we run a detailed exploration of the problem space and investigate an array of models ranging from purely textual to graph based to finally semi-supervised techniques using Graph Neural Networks (GNN) that utilize both textual and graph-based features. We run exhaustive experiments on two datasets -- Gab, which is loosely moderated and Twitter, which is strictly moderated. Overall the AGNN model achieves 0.791 macro F1-score on the Gab dataset and 0.780 macro F1-score on the Twitter dataset using only 5% of the labeled instances, considerably outperforming all the other models including the fully supervised ones. We perform detailed error analysis on the best performing text and graph based models and observe that hateful users have unique network neighborhood signatures and the AGNN model benefits by paying attention to these signatures. This property, as we observe, also allows the model to generalize well across platforms in a zero-shot setting. Lastly, we utilize the best performing GNN model to analyze the evolution of hateful users and their targets over time in Gab.

Punyajoy Saha, Binny Mathew, Kiran Garimella et al.

WhatsApp is the most popular messaging app in the world. Due to its popularity, WhatsApp has become a powerful and cheap tool for political campaigning being widely used during the 2019 Indian general election, where it was used to connect to the voters on a large scale. Along with the campaigning, there have been reports that WhatsApp has also become a breeding ground for harmful speech against various protected groups and religious minorities. Many such messages attempt to instil fear among the population about a specific (minority) community. According to research on inter-group conflict, such `fear speech' messages could have a lasting impact and might lead to real offline violence. In this paper, we perform the first large scale study on fear speech across thousands of public WhatsApp groups discussing politics in India. We curate a new dataset and try to characterize fear speech from this dataset. We observe that users writing fear speech messages use various events and symbols to create the illusion of fear among the reader about a target community. We build models to classify fear speech and observe that current state-of-the-art NLP models do not perform well at this task. Fear speech messages tend to spread faster and could potentially go undetected by classifiers built to detect traditional toxic speech due to their low toxic nature. Finally, using a novel methodology to target users with Facebook ads, we conduct a survey among the users of these WhatsApp groups to understand the types of users who consume and share fear speech. We believe that this work opens up new research questions that are very different from tackling hate speech which the research community has been traditionally involved in.

Binny Mathew, Sandipan Sikdar, Florian Lemmerich et al.

We introduce POLAR - a framework that adds interpretability to pre-trained word embeddings via the adoption of semantic differentials. Semantic differentials are a psychometric construct for measuring the semantics of a word by analysing its position on a scale between two polar opposites (e.g., cold -- hot, soft -- hard). The core idea of our approach is to transform existing, pre-trained word embeddings via semantic differentials to a new "polar" space with interpretable dimensions defined by such polar opposites. Our framework also allows for selecting the most discriminative dimensions from a set of polar dimensions provided by an oracle, i.e., an external source. We demonstrate the effectiveness of our framework by deploying it to various downstream tasks, in which our interpretable word embeddings achieve a performance that is comparable to the original word embeddings. We also show that the interpretable dimensions selected by our framework align with human judgement. Together, these results demonstrate that interpretability can be added to word embeddings without compromising performance. Our work is relevant for researchers and engineers interested in interpreting pre-trained word embeddings.

Binny Mathew, Anurag Illendula, Punyajoy Saha et al.

With the ongoing debate on 'freedom of speech' vs. 'hate speech' there is an urgent need to carefully understand the consequences of the inevitable culmination of the two, i.e., 'freedom of hate speech' over time. An ideal scenario to understand this would be to observe the effects of hate speech in an (almost) unrestricted environment. Hence, we perform the first temporal analysis of hate speech on Gab.com, a social media site with very loose moderation policy. We first generate temporal snapshots of Gab from millions of posts and users. Using these temporal snapshots, we compute an activity vector based on DeGroot model to identify hateful users. The amount of hate speech in Gab is steadily increasing and the new users are becoming hateful at an increased and faster rate. Further, our analysis analysis reveals that the hate users are occupying the prominent positions in the Gab network. Also, the language used by the community as a whole seem to correlate more with that of the hateful users as compared to the non-hateful ones. We discuss how, many crucial design questions in CSCW open up from our work.

Binny Mathew, Suman Kalyan Maity, Pawan Goyal et al.

Quora is a popular Q&A site which provides users with the ability to tag questions with multiple relevant topics which helps to attract quality answers. These topics are not predefined but user-defined conventions and it is not so rare to have multiple such conventions present in the Quora ecosystem describing exactly the same concept. In almost all such cases, users (or Quora moderators) manually merge the topic pair into one of the either topics, thus selecting one of the competing conventions. An important application for the site therefore is to identify such competing conventions early enough that should merge in future. In this paper, we propose a two-step approach that uniquely combines the anomaly detection and the supervised classification frameworks to predict whether two topics from among millions of topic pairs are indeed competing conventions, and should merge, achieving an F-score of 0.711. We also develop a model to predict the direction of the topic merge, i.e., the winning convention, achieving an F-score of 0.898. Our system is also able to predict ~ 25% of the correct case of merges within the first month of the merge and ~ 40% of the cases within a year. This is an encouraging result since Quora users on average take 936 days to identify such a correct merge. Human judgment experiments show that our system is able to predict almost all the correct cases that humans can predict plus 37.24% correct cases which the humans are not able to identify at all.

Binny Mathew, Ritam Dutt, Suman Kalyan Maity et al.

Anonymity forms an integral and important part of our digital life. It enables us to express our true selves without the fear of judgment. In this paper, we investigate the different aspects of anonymity in the social Q&A site Quora. The choice of Quora is motivated by the fact that this is one of the rare social Q&A sites that allow users to explicitly post anonymous questions and such activity in this forum has become normative rather than a taboo. Through an analysis of 5.1 million questions, we observe that at a global scale almost no difference manifests between the linguistic structure of the anonymous and the non-anonymous questions. We find that topical mixing at the global scale to be the primary reason for the absence. However, the differences start to feature once we "deep dive" and (topically) cluster the questions and compare the clusters that have high volumes of anonymous questions with those that have low volumes of anonymous questions. In particular, we observe that the choice to post the question as anonymous is dependent on the user's perception of anonymity and they often choose to speak about depression, anxiety, social ties and personal issues under the guise of anonymity. We further perform personality trait analysis and observe that the anonymous group of users has positive correlation with extraversion, agreeableness, and negative correlation with openness. Subsequently, to gain further insights, we build an anonymity grid to identify the differences in the perception on anonymity of the user posting the question and the community of users answering it. We also look into the first response time of the questions and observe that it is lowest for topics which talk about personal and sensitive issues, which hints toward a higher degree of community support and user engagement.

Binny Mathew, Suman Kalyan Maity, Pratip Sarkar et al.

Word senses are not static and may have temporal, spatial or corpus-specific scopes. Identifying such scopes might benefit the existing WSD systems largely. In this paper, while studying corpus specific word senses, we adapt three existing predominant and novel-sense discovery algorithms to identify these corpus-specific senses. We make use of text data available in the form of millions of digitized books and newspaper archives as two different sources of corpora and propose automated methods to identify corpus-specific word senses at various time points. We conduct an extensive and thorough human judgment experiment to rigorously evaluate and compare the performance of these approaches. Post adaptation, the output of the three algorithms are in the same format and the accuracy results are also comparable, with roughly 45-60% of the reported corpus-specific senses being judged as genuine.