List of Standard Dimensions

This page provides a comprehensive overview of the standard dimensions available for analyzing conversational data, organized by their scope of analysis.

Session Analysis Dimensions

The following dimensions provide insights at the session level, analyzing the entire conversation transcript:

Dimension Name	Description
Reason	A concise label describing the main topic or intent of the full session transcript
Category	A business-level category describing the reason or primary driver for the session
Effort	A score from 1-5 predicting the level of difficulty for the user to complete a task, 5 being the most effort (conversational)
Helpfulness	A score from 1-5 predicting the helpfulness of the agent assisting the user, 5 being the most helpful (conversational)
Predicted Feedback	Simulated user feedback summarizing satisfaction with a request or experience
Feedback Cluster	Predicted feedback from simulated users clustered according to similarity
Predicted Solution	A proposed improvement or fix inferred to enhance user experience in the future
Solution Cluster	Predicted solutions clustered by similar improvement themes
Agent Sentiment	The overall tonality or emotion of the outgoing user
Agent Sentiment Cluster	Outgoing user sentiment clustered by positive, neutral, negative
User Sentiment	The overall tonality or emotion of the incoming user
User Sentiment Cluster	Incoming user sentiment clustered by positive, neutral, negative
Outcome	The predicted end state of the interaction: complete, abandoned, escalated (conversational)
Predicted Rating	A score from 1-10 predicting the user's satisfaction rating
Predicted Product	The product name or service mentioned, if any
Recommendation	A recommendation to improve a product or service made by the user
Recommendation Cluster	User recommendations to improve a product or service clustered by similarity
Predicted Success Score	A score between 1 and 100 estimating the percentage of the request handled successfully (conversational)
Summary	A short but detailed summary of the full session transcript
Summary Cluster	Summaries of the full session transcript clustered by similarity

Section Analysis Dimensions

The following dimensions provide insights at the section level, analyzing specific portions of conversation transcripts:

Dimension Name	Description
Reason	A concise label describing the main topic or intent of a section of the transcript
Category	A grouping of section-level reasons clustered by similar underlying themes
User Sentiment	A label describing the tone or emotion of incoming user messages in a section (conversational)
User Sentiment Cluster	A grouping of section-level incoming user sentiment clustered by positive, neutral, or negative (conversational)
Agent Sentiment	A label describing the tone or emotion of outgoing user messages in a section (conversational)
Agent Sentiment Cluster	A grouping of section-level outgoing user sentiment clustered by positive, neutral, negative (conversational)
User Reason	The primary topic or intent expressed by the incoming user in a section of the transcript (conversational)
User Explanation	A short description explaining the user's reason label for a section (conversational)
Agent Reason	The main topic or intent expressed by the outgoing user in a section of the transcript (conversational)
Agent Explanation	A short description explaining the outgoing user's reason label for a section (conversational)
Predicted Intent	A suggested automation intent inferred from a section of the transcript (conversational)
Product	The product or service mentioned in a section of the transcript, if any
Summary	A concise summary of the content and purpose of a section
Repeat	True if the section repeats or revisits an earlier topic within the transcript
Specificity Score	A score from 1-5 reflecting the level of detail and actionability of content in a section

Note: Dimensions marked as "(conversational)" are specifically designed for analyzing conversational data and dialogue interactions.