iiinova — where your digital self keeps evolving

Rohan @rohandas.net · 11/05/2026

At the domain level, immigration is driven by character portrayals (who is being portrayed and how) while gun control is driven by judicial conflict and institutional enforcement schemas (what is happening), suggesting the two domains are organized around different narrative dimensions. 🧵9/10

We find that for immigration, characters play a central role in driving frame prediction, particularly immigrants. In contrast, gun control frame predictions are driven by narrative clusters, particularly those dealing with judicial and legal schemas.

120

Rohan @rohandas.net · 11/05/2026

At the frame level, we found narrative signatures vary meaningfully across policy frames. For example, the Legality/Constitutionality frame in gun control is dominated by judicial conflict schemas, suggesting legal framing emphasizes courtroom battles while political actors remain peripheral. 🧵8/10

Legal framing of gun control coverage tends to emphasize courtroom battles rather than legislative politics.

110

Rohan @rohandas.net · 11/05/2026

The SHAP analysis operates at three levels: instance, frame, and domain. At the instance level, narrative schema features capture equivalent information to RoBERTa embeddings, while offering more succinct and interpretable representations. 🧵7/10

SHAP feature importance comparison between RoBERTa embeddings and narrative features for frame prediction on a single test instance, showing both approaches capture similar semantic patterns.

110

Rohan @rohandas.net · 11/05/2026

We evaluated 4,000 news articles across immigration and gun control. Schema quality was validated by a trained linguist, with 94-96% rated high quality, consistently capturing Entman's framing elements. We also conducted a SHAP analysis using induced schemas as features for frame prediction. 🧵5/10

Examples of high-quality generated schemas, evaluated on their overall coherence, as well as on striking a balance between coverage of and specificity to the sentences included in the context.

120

Rohan @rohandas.net · 11/05/2026

We present a framework for unsupervised, domain-agnostic narrative schema induction that scales to any large corpora. Our method extracts causal event chains, assigns character roles (Hero/Threat/Victim) to entities, and uses these as constraints in a structured clustering framework. 🧵3/10

For a large scale news corpus, we first construct narrative event chains and obtain character and role annotations for them. We then cluster these narrative chains using the character and role information as constraints. The generated narrative clusters are representative of fine-grained and nuanced narrative schemas.

110

Rohan @rohandas.net · 11/05/2026

Existing NLP approaches often build narratives bottom-up from extractable atomic units like predicate argument structures or entities. While highly scalable, these methods seldom capture the evaluative and ideological dimensions central to how meaning is constructed in the media. 🧵2/10

Law enforcement spending framed as worker protection versus government waste across policy domains.

120

Rohan @rohandas.net · 11/05/2026

Computational approaches to media narrative analysis either miss nuanced storytelling patterns through coarse-grained analysis, or require domain-specific taxonomies that limit scalability. We show joint event and character modeling can address this gap. Details in our #ACL2026 (Main) paper. 🧵1/10

Paper Title: A Structured Clustering Approach for Inducing Media Narratives

Authors: Rohan Das, Advait Deshmukh, Alexandria Leto, Zohar Naaman, I-Ta Lee, Maria Leonor Pacheco

12411

Rohan @rohandas.net · 10/05/2026

Our contributions: 1. A 3-pronged evaluation framework for interactive qualitative coding systems. 2. Experimental results comparing synchronous and asynchronous coding. 🧵3/7

In this study, we measure the quality of coded themes using different interactive systems under different coding configurations.

110

Rohan @rohandas.net · 10/05/2026

We tasked researchers to use NLP tools to identify themes from a dataset of 5.5k Facebook ads on climate change. Here are two of the resulting code-books. Do you think one code-book is better than the other? 🧵2/7

Themes as they belong to two different qualitative codebooks derived from the same dataset.

100

Rohan @rohandas.net · 10/05/2026

What’s the best way to analyze online discourse on any given topic? Is there a right way to use NLP tools to sift through massive datasets? To find out, we tested several tools across different collaboration settings and report findings in an #ACL2026 (Main) paper: arxiv.org/abs/2408.09030 🧵1/7

Paper Title: Effects of Collaboration on the Performance of Interactive Theme Discovery Systems

Authors: Alvin Po-Chun Chen, Rohan Das, Dananjay Srinivas, Alexandra Barry, Maksim Seniw, Maria Leonor Pacheco

1125

Rohan @rohandas.net · 05/05/2026

We evaluated 3 NLP-assisted coding systems (topic model, relational, LLM-based) under synchronous vs. asynchronous collaboration, on 2 corpora: 85K COVID vaccine tweets and 5.5K climate change ads. The study involved 33 researchers across 2 universities, spanning 30 coding experiments. 🧵2/5

110

Rohan @rohandas.net · 19/11/2024

Bonus: You get to work out of a lab with not only windows but also gorgeous views of the Rockies! ⛰️

010