N.V.J.K KartikKartik

Book chapter · IIT Bombay · Published online 23 Oct 2025 · T4E 2024 · Springer Lecture Notes in Educational Technology

Triggers of shared regulation

When a group stops and asks “wait, is this working?”, what set it off? Could a model spot that moment in the conversation?

Plate: Triggers of shared regulation.

What I did

First author, with seven co-authors from the same research group.

I applied Conditional Random Fields to detect the triggers of socially shared regulation in group problem-solving.

Springer ↗DOI 10.1007/978-981-95-1734-3_6 ↗

Authors

N.V.J.K Kartik, Priyesh Gupta, Vinayak, Aarsh Desai, Vishwas Badhe, T S Ashwin, Manjunath Vanahalli, Ramkumar Rajendran

Full title

Unlocking the Triggers: Automating the Identification of Triggers of Socially Shared Metacognitive Regulation in Collaborative Problem-Solving

In plain words

Socially shared metacognitive regulation (SSMR) is a group jointly checking and steering its own thinking. “Hold on, are we even solving the right problem?” is a small example.

Researchers usually find these moments by hand-coding transcripts, which is slow. This paper works on detecting what triggers them automatically, from the group’s conversation.

How it works

  • The approach treats the conversation as a sequence and labels trigger moments. It uses Conditional Random Fields over collaborative problem-solving discourse.
  • Published in the T4E 2024 proceedings (Vol. 2), pp. 43–51.

What it established

  • It frames trigger detection as an automatable sequence-labelling task, a step toward analysing collaboration at a scale hand-coding can’t reach.
Where it stops
  • It works from what people said. Tone, posture and gaze are not in the model. Later work from the group proposes a multimodal extension.
  • I haven’t reproduced the paper’s headline numbers here. Read the chapter for them.

Receipts

  1. Springer — T4E 2024 Proceedings Vol. 2, pp. 43–51Title, author order (Kartik first), series, pages, published online 23 Oct 2025 (via Crossref metadata).
  2. Crossref metadata for DOI 10.1007/978-981-95-1734-3_6Author list, online publication date, page range.
  3. Previous portfolio, as it stood before this redesign (self-reported)Self-reported: IIIT Dharwad (Data Science & AI), research at IIT Bombay Educational Technology, CRF approach for SSMR.
  4. Badhe & Rajendran — multimodal follow-up on SSMR triggers (Springer, 2026)The group later proposed extending verbal-only trigger detection with video (posture, head movement).

Every source on the site

Next, a little further back in timeAgentCompass →