RealTalk Themes

Discover conversation themes via clustering, relate those to human generated highlights, and explore temporal patterns in how conversations develop.

Using natural language processing and machine learning, we discover emergent themes and patterns within our dialogue data set. Exploring how those relate to human-generated themes and patterns, we can identify how various means of sensemaking relate to one another. Further, by analyzing the dialogue as a whole, we are able to identify relationships between themes temporally and visualize how conversations unfold overtime to help interpret and understand conversations and communities.

Program: Scaffolding
Phase: Early Exploration
Tags: Conversation Visualization, NLP
Back to Projects