Educational videos are widely used, but accessibility guidelines beyond captions for d/Deaf and Hard-of-Hearing (DHH) learners remain limited. Mayer’s multimedia learning theory assumes visual-auditory dual-channel processing, yet DHH learners with limited access to the auditory channel have distinct visual abilities and cognitive demands. This paper introduces motion-driven design ideas to support cognitive processing and improve video-based learning for DHH learners. Through a three-phase study, we identified four key challenges—such as misaligned content and visual overload—and proposed four design ideas that extend multimedia learning theory. We then evaluated these ideas with 16 DHH learners and 6 experts in Deaf education. The results show that motion-driven approaches reduce misalignment, ease visual attention switching, and improve the integration of visual and textual information across video types. For example, guiding visual attention switching minimizes confusion in complex visual contexts, such as programming demonstrations, while using relevant visuals enriches talking-head videos with graphics to clarify abstract ideas in captions. More research is needed to develop these promising ideas into well-defined principles.

The rise of large language models (LLMs) has created a need for advanced benchmarking systems beyond traditional setups. To this end, we introduce QUENCH, a novel text-based English Quizzing Benchmark manually curated and transcribed from YouTube quiz videos. QUENCH possesses masked entities and rationales for the LLMs to predict via generation. At the intersection of world knowledge, geographical context, and common sense reasoning, QUENCH helps assess world knowledge and deduction capabilities of LLMs via a zero-shot, open-domain quizzing setup. We perform an extensive evaluation on 7 LLMs and 4 metrics, investigating the influence of model size, prompting style, geographical context, and gold-labeled rationale generation. The benchmarking concludes with an error analysis of various types of generative errors to which the LLMs are prone.

For subjective tasks such as hate detection, where people perceive hate differently, the Large Language Model's (LLM) ability to represent diverse groups is unclear.By including additional context in prompts, we comprehensively analyze LLM's sensitivity to geographical priming, persona attributes, and numerical information to assess how well the needs of various groups are reflected.Our findings on two LLMs, five languages, and six datasets reveal that mimicking persona-based attributes leads to annotation variability.Meanwhile, incorporating geographical signals leads to better regional alignment.We also find that the LLMs are sensitive to numerical anchors, indicating the ability to leverage community-based flagging efforts and exposure to adversaries.Our work provides preliminary guidelines and highlights the nuances of applying LLMs in culturally sensitive cases. 1

Social media is awash with hateful content, much of which is often veiled with linguistic and topical diversity. The benchmark datasets used for hate speech detection do not account for such divagation as they are predominantly compiled using hate lexicons. However, capturing hate signals becomes challenging in neutrally-seeded malicious content. Thus, designing models and datasets that mimic the real-world variability of hate warrants further investigation.

With an increasing outreach of digital platforms in our lives, researchers have taken a keen interest in studying different facets of social interactions. Analyzing the spread of information ( aka diffusion) has brought forth multiple research areas such as modelling user engagement, determining emerging topics, forecasting the virality of online posts and predicting information cascades. Despite such ever-increasing interest, there remains a vacuum among easy-to-use interfaces for large-scale visualization of diffusion models. In this article, we introduce DiVA — Di ffusion V isualization and A nalysis, a tool that provides a scalable web interface and extendable APIs to analyze various diffusion trends on networks. DiVA uniquely offers support for simultaneous comparison of two competing diffusion models and even the comparison with the ground-truth results, which help develop a coherent understanding of real-world scenarios. Along with performing an exhaustive feature comparison and system evaluation of DiVA against publicly-available web interfaces for information diffusion, we conducted a user study to understand the strengths and limitations of DiVA . We noticed that evaluators had a seamless user experience, especially when analyzing diffusion on large networks.

Curbing online hate speech has become the need of the hour; however, a blanket ban on such activities is infeasible for several geopolitical and cultural reasons. To reduce the severity of the problem, in this paper, we introduce a novel task, hate speech normalization, that aims to weaken the intensity of hatred exhibited by an online post. The intention of hate speech normalization is not to support hate but instead to provide the users with a stepping stone towards non-hate while giving online platforms more time to monitor any improvement in the user's behavior. To this end, we manually curated a parallel corpus - hate texts and their normalized counterparts (a normalized text is less hateful and more benign). We introduce NACL, a simple yet efficient hate speech normalization model that operates in three stages - first, it measures the hate intensity of the original sample; second, it identifies the hate span(s) within it; and finally, it reduces hate intensity by paraphrasing the hate spans. We perform extensive experiments to measure the efficacy of NACL via three-way evaluation (intrinsic, extrinsic, and human-study). We observe that NACL outperforms six baselines - NACL yields a score of 0.1365 RMSE for the intensity prediction, 0.622 F1-score in the span identification, and 82.27 BLEU and 80.05 perplexity for the normalized text generation. We further show the generalizability of NACL across other platforms (Reddit, Facebook, Gab). An interactive prototype of NACL was put together for the user study. Further, the tool is being deployed in a real-world setting at Wipro AI as a part of its mission to tackle harmful content on online platforms.

While the rise in popularity of social media is seen as a hugely positive development, it is also accompanied by a proliferation of hate speech, which has recently become a major concern. On the one hand, hateful content creates an unsafe environment for certain members of society. On the other hand, manual moderation causes distress to content moderators, and the volume of harmful content is far beyond what human moderators can manually flag and react to. Thus, researchers in machine learning, social computing, and other areas have worked on developing tools to help automate the process. While initially studied as a text classification problem, over time, researchers realized that hate speech is multi-faceted and requires analysis of the role of linguistic expressions, context, and network structure, while using inspiration from psychology and user behavior, among others. With this in mind, we provide a holistic view of what the research community has explored so far, and what we believe are promising future research directions.

Online hate speech, particularly over microblogging platforms like Twitter, has emerged as arguably the most severe issue of the past decade. Several countries have reported a steep rise in hate crimes infuriated by malicious hate campaigns. While the detection of hate speech is one of the emerging research areas, the generation and spread of topic-dependent hate in the information network remain under-explored. In this work, we focus on exploring user behavior, which triggers the genesis of hate speech on Twitter and how it diffuses via retweets. We crawl a large-scale dataset of tweets, retweets, user activity history, and follower networks, comprising over 161 million tweets from more than 41 million unique users. We also collect over 600k contemporary news articles published online. We characterize different signals of information that govern these dynamics. Our analyses differentiate the diffusion dynamics in the presence of hate from usual information diffusion. This motivates us to formulate the modeling problem in a topic-aware setting with real-world knowledge. For predicting the initiation of hate speech for any given hashtag, we propose multiple feature-rich models, with the best performing one achieving a macro F1 score of 0.65. Meanwhile, to predict the retweet dynamics on Twitter, we propose RETINA, a novel neural architecture that incorporates exogenous influence using scaled dot-product attention. RETINA achieves a macro F1-score of 0.85, outperforming multiple state-of-the-art models. Our analysis reveals the superlative power of RETINA to predict the retweet dynamics of hateful content compared to the existing diffusion models.

Modeling user engagement dynamics on social media has compelling applications in market trend analysis, user-persona detection, and political discourse mining. Most existing approaches depend heavily on knowledge of the underlying user network. However, a large number of discussions happen on platforms that either lack any reliable social network (news portal, blogs, Buzzfeed) or reveal only partially the inter-user ties (Reddit, Stackoverflow). Many approaches require observing a discussion for some considerable period before they can make useful predictions. In real-time streaming scenarios, observations incur costs. Lastly, most models do not capture complex interactions between exogenous events (such as news articles published externally) and in-network effects (such as follow-up discussions on Reddit) to determine engagement levels. To address the three limitations noted above, we propose a novel framework, ChatterNet, which, to our knowledge, is the first that can model and predict user engagement without considering the underlying user network. Given streams of timestamped news articles and discussions, the task is to observe the streams for a short period leading up to a time horizon, then predict chatter: the volume of discussions through a specified period after the horizon. ChatterNet processes text from news and discussions using a novel time-evolving recurrent network architecture that captures both temporal properties within news and discussions, as well as influence of news on discussions. We report on extensive experiments using a two-month-long discussion corpus of Reddit, and a contemporaneous corpus of online news articles from the Common Crawl. ChatterNet shows considerable improvements beyond recent state-of-the-art models of engagement prediction. Detailed studies controlling observation and prediction windows, over 43 different subreddits, yield further useful insights.

Many real-world problems, such as placement of surveillance cameras and pricing of hotel rooms with a view, require the ability to determine the visibility of a given target object from different locations. Advances in large-scale 3D modeling (e.g., 3D virtual cities) provide us with data that can be used to solve these problems with high accuracy. In this paper, we investigate the problem of finding the location which provides the best view of a target object with visual obstacles in 2D or 3D space, for example, finding the location that provides the best view of fireworks in a city with tall buildings. To solve this problem, we first define the quality measure of a view (i.e., visibility measure) as the visible angular size of the target object. Then, we propose a new query type called the k-Maximum Visibility (kMV) query, which finds k locations from a set of locations that maximize the visibility of the target object. Our objective in this paper is to design a query solution which is capable of handling large-scale city models. This objective precludes the use of approaches that rely on constructing a visibility graph of the entire data space. As a result, we propose three approaches that incrementally consider relevant obstacles in order to determine the visibility of a target object from a given set of locations. These approaches differ in the order of obstacle retrieval, namely: query centric distance based, query centric visible region based, and target centric distance based approaches. We have conducted an extensive experimental study on real 2D and 3D datasets to demonstrate the efficiency and effectiveness of our solutions.