top of page

TEAM

Rhetorics for Attitudinal Convergence

Adam Perhala, Jungbae An

clear.png

We identify rhetorical strategies that can induce attitudinal convergence in discussions. We will track attitudinal changes in longitudinal discussions and infer the factors driving the convergence of attitudes among discussion participants:

(1) Classification: Our first task is to classify the attitudes of discussion participants towards an issue at a given time. [I'm thinking of LDA. But, I'd use a DNN-based classification algorithm (e.g., VAE) or an application of LLM API, if anyone is familiar with either.]
(2) Convergence: We may examine similarity/modularity of participants' attitudes.
(3) Network Inference: Finally, we will infer the cause of converging attitudes from participant, speech, and issue characteristics.

* Data: The standing committee transcripts of the US Senate, a time-stamped text dataset are available online. But, I'd prefer a more industry-friendly data, if available.

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL
bottom of page