Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2611.06226

Understanding the Perception of Transliterated Offensive Social Media Posts Among Banglish and Hinglish Users

Dhiman Goswami·Zhicong Lu·Sanchari Das
View original source ↗Read source document ↗Export metadata ↓

Abstract

Effective content moderation depends on understanding not just what harmful content looks like, but how users perceive and respond to it. When offensive language is written in transliterated forms such as Banglish or Hinglish (Bangla or Hindi in Latin script), it may be interpreted very differently than in native scripts, with direct consequences for whether users report, ignore, or recognize it as harmful. Yet little is known about how script choice relates to user perception, or how these perceptions vary across demographic groups. To investigate these perceptions, we conducted a user study with $310$ participants ($193$ Hindi, $60$ Bangla, $57$ Bangla-Hindi bilingual speakers) who evaluated offensive posts in native scripts and transliterations. Participants provided offensiveness scores ($4.03$ for Banglish, $3.86$ for Hinglish) to transliterated posts, yet reporting fell from $20.9\%$ (native - Bangla, Hindi) to $18.2\%$ (transliterated - Banglish, Hinglish). Ignoring remained the dominant response (range $25.38\%-44\%$ overall). Women reported at higher probability than men for native-script ($23.6\%$ vs. $17.7\%$) and showed almost identical reporting probability in comparison to men for transliterated posts ($17.5\%$ vs. $17.9\%$), and younger participants ($<40$) rated Banglish content as more offensive than older participants ($\geq 40$), (Overall offensive score: $3.99$ vs. $3.78$). These findings identify transliteration as critical blind spot for moderation, underscoring the need for linguistically sensitive approaches to online safety.

Explore related subjects

cs.HC

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Dhiman Goswami, Zhicong Lu, Sanchari Das. 2026-11-04. Understanding the Perception of Transliterated Offensive Social Media Posts Among Banglish and Hinglish Users. https://arxiv.org/abs/2611.06226

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

HarmonICA: Quasi-Time-Invariant ICA for Motor Unit Decomposition during Dynamic Contractions

Decomposing surface electromyography (EMG) into the spike trains of individual motor neurons is a long-standing inverse problem and a key step toward motor-neuron-driven neural interfaces such as prosthetics and exoskeletons. The standard approach, independent component analysis (ICA) of the multichannel signal, assumes that the mixing from neurons to electrodes is stationary in time. This assumption fails during movement, when volume-conductor deformation makes the mixing time-varying, and current decomposition algorithms are correspondingly restricted to isometric contractions. We introduce HarmonICA, a quasi-time-invariant ICA formulation in which a static linear separator is preceded by a learned, low-rank, time-varying invertible transformation. The separator is trained with an independence loss on the uncompensated projection, and the transformation with a stationarity loss on the recovered source. Gradients are not shared between the two, so source extraction remains a classical linear-ICA problem while non-stationary distortion is absorbed by the transformation. The closed-form inverse of the transformation enables per-spike subtraction with a time-varying template during sequential peel-off. On a public simulated benchmark of dynamic high-density EMG with ground-truth spike trains, the method outperforms five adaptive ICA baselines at every recall threshold, recovering more units at a higher accuracy.

cs.HC↗
arXiv2026

A State-of-the-Art Review of Computational Models for Analyzing Longitudinal Wearable Sensor Data in Healthcare

Wearable devices are increasingly used as tools for biomedical research, as the continuous stream of behavioral and physiological data they collect can provide insights about our health in everyday contexts. Long-term tracking, defined in the timescale of months of year, can provide insights of patterns and changes as indicators of health changes. These insights can make medicine and healthcare more predictive, preventive, personalized, and participative (The 4P's). However, the challenges in modeling, understanding and processing longitudinal data are a significant barrier to their adoption in research studies and clinical settings. In this paper, we review and discuss three models used to make sense of longitudinal data: routines, rhythms and stability metrics. We present the challenges associated with the processing and analysis of longitudinal wearable sensor data, with a special focus on how to handle the different temporal dynamics at various granularities. We then discuss current limitations and identify directions for future work. This review is essential to the advancement of computational modeling and analysis of longitudinal sensor data for pervasive healthcare.

cs.HC↗
arXiv2026

Does Explanation Correctness Matter? Linking Computational XAI Evaluation to Human Understanding

Explainable AI (XAI) methods are commonly evaluated using functional correctness metrics, also called faithfulness or fidelity, which estimate how closely an explanation reflects the model's reasoning. Higher correctness is assumed to produce better human understanding, but this link has not been tested with controlled levels. We conducted a user study (N=200) that manipulated explanation correctness at four levels (100%, 85%, 70%, 55%) in a synthetic time series classification task where participants could not rely on domain knowledge or visual intuition. Explanations were constructed to resemble feature attributions, with correctness defined against a known ground truth rather than estimated from a trained model. Participants predicted a simulated AI's decisions from these explanations (forward simulation), and we used their forward simulation accuracy as a proxy for understanding. Correctness affected forward simulation accuracy, but not at every level: accuracy dropped at 70% and 55% relative to 100%, with no further reduction below 70% and no conclusive difference between 85% and 100%. Lower correctness reduced how many participants predicted accurately rather than lowering accuracy uniformly, and even with fully correct explanations, only a subset of participants predicted well above chance. In exploratory analyses, some participants predicted decisions accurately but described the wrong pattern when asked what the AI used, and self-reported ratings correlated with forward simulation accuracy only when explanations were fully correct. These findings suggest that, for feature-attribution-style explanations, not all differences in functional correctness translate to differences in human understanding, highlighting the need to validate functional metrics against human outcomes.

cs.HC↗