Here's what you need to know:
Traditionally, sentiment scores have been limited by simple, sentence-based models that don't truly capture the dynamics of a conversation. These models only reflect the type of words used, like "good", "bad" or "sorry," without considering the overall context.
Dubber's Conversation Sentiment score uses state-of-the-art language models to analyse entire conversations. It considers speaking turns and full context, providing a more accurate picture of how sentiment evolves throughout an interaction.
We've thoroughly tested this new approach using internal datasets and subjective user evaluations. Results confirm that our new scores more accurately reflect speaker sentiment in real conversations.
Why This Matters:
This upgrade positions Dubber at the forefront of Conversation Intelligence. While others rely on simplistic word-based sentiment scores, our solution offers a deeper, more contextual understanding of conversation dynamics. This advancement will help our users:
- Better analyse and understand customer interactions.
- Set up AI-powered alerts for negative sentiment calls, enabling rapid response to critical situations.
- Easily filter and review calls with specific sentiment patterns for quality assurance and training.
- Dubber Insights users can uncover valuable patterns by using Sentiment with Dubber Moments, such as identifying Complaint calls with negative sentiment or analysing Sales Close Calls with unexpected sentiment shifts.
Availability and Access:
Please note that the new sentiment model is available exclusively for users with specific plans. You can access this enhanced feature if you have a Dubber Insight SKU, UC+1, or any SKU enabled for Dubber Insights.
When activated, the new sentiment analysis will be accessible on both portal versions. For users with AI Plans not enabled for Insights (without the new portal experience), the previous sentiment model will continue to be used.