Welcome to Deliberation Integrity Monitor

The Deliberation Integrity Monitor (DIM) is a tool for analyzing the quality of deliberative discussions. It helps you understand participation patterns, identify imbalances, and generate comprehensive reports.

Features

  • Upload & Match: Upload discussion transcripts and optionally match participants with demographic data
  • Inclusion Analysis: Analyze speaking time and word count by participant or demographic groups
  • Reflection Analysis: Evaluate deliberation quality through argumentation and reasoning patterns
  • Sentiment Analysis: Detect emotions in the discussion using a multilingual lexicon, with optional AI interpretation
  • Topic Analysis: Discover recurring themes using topic modeling, with optional AI interpretation
  • Generate Report: Create comprehensive PDF reports with AI-enhanced interpretations

Getting Started

  1. Upload a transcript in VTT format (see Upload Data tab)
  2. Optionally upload demographic data about participants
  3. Match speakers to demographics
  4. Analyze participation patterns in the Inclusion tab
  5. Generate a comprehensive PDF report

License Key

PRO Features: Some advanced features require a valid license key.

To request a license key, please email the system administrator at:

admin@deliberation-monitor.org




Free vs PRO

Feature Free PRO
Upload transcripts Up to 10 participants Unlimited
Inclusion analysis ✓ ✓
Reflection analysis ✓ ✓
Topic analysis ✓ ✓
Sentiment analysis ✓ ✓
AI topic interpretation — ✓
AI-enhanced reports — ✓

1. Upload Transcript

Upload a transcript file in VTT format with speaker names and timestamps.

Click here to download a sample transcript.

2. Upload Demographics (Optional)

Upload demographic data about participants to enable group-based analysis.

Required column: name (participant label — can be anonymous, e.g. P1, P2, P3)

Optional column: speaker_id — when present, speakers are matched automatically using this ID (e.g. S01, S02). Useful with noScribe transcripts where speakers are labelled S00, S01, etc.

Demographic columns: Use snake_case format (e.g., gender , education_attainment , age , political_affiliation )

Click here to download a sample demographics file.

Preview Data


Inclusion

Inclusion is measured via speaking time. Speaking time indicates the extent to which participants contributed to the discussion.

Reflection

This analysis examines the quality of deliberative reasoning by identifying linguistic markers of reflective thinking. Each measure is calculated by counting specific patterns in the transcript. Hover over each statistic below to see how it is calculated.

Sentiment

Sentiment analysis detects emotions expressed in each utterance using a lexicon-based method. Eight emotions are measured — Joy, Trust, Anticipation, Surprise, Sadness, Fear, Disgust, and Anger — by matching utterances against curated word lists for English, German, French, and Italian.

Topics

Identify recurring themes in the discussion using topic modeling. Topics are discovered automatically from the transcript text and represent clusters of words that tend to appear together.

Generate Report

Generate a comprehensive PDF report of your deliberation analysis.

Interpretations for Report

Interpretations from the analysis tabs are pre-loaded below. Edit, generate with AI (PRO license required), or leave blank to omit a section from the report.