The interactivity score measures how engaged and responsive each speaker is in a conversation. Instead of only tracking how much someone talks, it highlights how well they participate in dialogue through questions, answers, and back-and-forth exchanges.
How the score is calculated
For each speaker, the system tracks:
Times spoken – number of times they spoke.
Total words – how many words they contributed.
Conversational turns – how often they participated in back-and-forth dialogue.
Questions asked – counted whenever a sentence ends with a “?”.
Questions answered – tracked when they speak immediately after someone else asks a question.
Each metric is normalized (calculated as a ratio of the total group activity), so contributions are compared fairly across all speakers.
Not all factors are equally important. To calculate the final score, we apply weights that reflect the value of each type of contribution:
10% → Message frequency
10% → Word contribution
30% → Conversational turns (highest weight – encourages dialogue)
25% → Questions asked (shows curiosity and initiative)
20% → Questions answered (shows responsiveness)
These weights add up to 100%, meaning each speaker’s final score is a blend of all five factors.
Let’s say a speaker contributes:
15% of all messages
20% of all words
30% of all turns
25% of all questions asked
10% of all questions answered
Their interactivity score would be:
(0.15 × 10%) + (0.20 × 10%) + (0.30 × 30%) + (0.25 × 25%) + (0.10 × 20%)
= 0.215 → 21.5%
Range: 0 to 1 (or 0% to 100%).
Higher score = more interactive participation.
The score emphasizes engagement (questions, answers, turns) more than just talking time (messages, words).
In short, someone who actively asks and answers questions in dialogue will score higher than someone who only speaks at length without interaction.
