CIG: Measuring Conversational Information Gain in Deliberative Dialogues with Semantic Memory Dynamics

2026-04-16

AI Summary

This study addresses the challenge of objectively quantifying informational progress in public deliberation beyond mere politeness or argumentative formality. It proposes the Conversational Information Gain (CIG) framework, which models information advancement as the dynamic evolution of semantic memory. By leveraging large language models to extract atomic claims and incrementally integrating them into a structured memory state, CIG evaluates each utterance’s contribution to collective understanding along three dimensions: novelty, relevance, and inferential breadth. Experiments on 80 real-world deliberative dialogues demonstrate that CIG—grounded in dynamic features such as memory updates—significantly outperforms traditional heuristic approaches, more accurately predicting human judgments of information gain and yielding a high-performing, interpretable evaluation model.

Technology Category

• Natural Language Processing: Discourse, Pragmatics & Argument Mining
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Measuring the quality of public deliberation requires evaluating not only civility or argument structure, but also the informational progress of a conversation. We introduce a framework for Conversational Information Gain (CIG) that evaluates each utterance in terms of how it advances collective understanding of the target topic. To operationalize CIG, we model an evolving semantic memory of the discussion: the system extracts atomic claims from utterances and incrementally consolidates them into a structured memory state. Using this memory, we score each utterance along three interpretable dimensions: Novelty, Relevance, and Implication Scope. We annotate 80 segments from two moderated deliberative settings (TV debates and community discussions) with these dimensions and show that memory-derived dynamics (e.g., the number of claim updates) correlate more strongly with human-perceived CIG than traditional heuristics such as utterance length or TF--IDF. We develop effective LLM-based CIG predictors paving the way for information-focused conversation quality analysis in dialogues and deliberative success.

Problem

• Conversational Information Gain
• Deliberative Dialogues
• Semantic Memory
• Information Progress
• Collective Understanding

Innovation

• Conversational Information Gain
• Semantic Memory Dynamics
• Atomic Claim Extraction
• Deliberative Dialogue Evaluation
• LLM-based Prediction