Prof. Ponnurangam Kumaraguru and his students Arvindh Arun, Karuna K Chandra, Akshit Sinha, Balakumar Velayutham and;  Jashn Arora and Manish Jain from  Google DeepMind, Bangalore received best paper award for their research work on  Topo Goes Political: TDA-Based Controversy Detection in Imbalanced Reddit Political Data at BeyondFacts – 5nd International Workshop on Knowledge Graphs for Online Discourse Analysis (KnOD 2025), collocated with The Web Conference 2025 held at Sydney, Australia from 28 April to 2 May. Here is the summary of the paper as explained by the authors:

The detection of controversial content in political discussions on the Internet is a critical challenge in maintaining healthy digital discourse. Unlike much of the existing literature that relies on synthetically balanced data, our work preserves the natural distribution of controversial and non-controversial posts. This real-world imbalance highlights a core challenge that needs to be addressed for practical deployment. Our study re-evaluates well-established methods for detecting controversial content. We curate our own dataset focusing on the Indian political context that preserves the natural distribution of controversial content, with only 12.9% of the posts in our dataset being controversial. This disparity reflects the true imbalance in real-world political discussions and highlights a critical limitation in the existing evaluation methods. Benchmarking on datasets that model data imbalance is vital for ensuring real-world applicability. Thus, in this work, (i) we release our dataset, with an emphasis on class imbalance, that focuses on the Indian political context, (ii) we evaluate existing methods from this domain on this dataset and demonstrate their limitations in the imbalanced setting, (iii) we introduce an intuitive metric to measure a model’s robustness to class imbalance, (iv) we also incorporate ideas from the domain of Topological Data Analysis, specifically Persistent Homology, to curate features that provide richer representations of the data. Furthermore, we benchmark models trained with topological features against established baselines.

Conference page: https://www2025.thewebconf.org/

June 2025