Pre-Chat Intent Prediction Using Candidate Question Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users often struggle to formulate questions for chatbots, leading to friction in interactions due to unknown intent or phrasing issues.
Innovation Solution
A system that predicts user intent through neural networks, presenting candidate questions on a chat interface for selection, and updates the model based on user feedback to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the chatbot waits for the user to formulate questions, then the user can ask their own questions, but the user experiences friction due to not knowing what to ask or how to phrase questions
Solution Approach 1:
The system performs preliminary action by predicting user intent and generating candidate questions before the user actually asks something. The chatbot analyzes user profile, conversation history, and contextual information to proactively suggest relevant questions, eliminating the need for the user to formulate questions from scratch and reducing interaction friction.
Solution Approach 2:
The system introduces an intermediary mechanism - the intent prediction model and candidate question generation - that mediates between the user's implicit information needs and the chatbot's response generation. This intermediary layer translates user context into actionable question suggestions, bridging the gap between user intent and chatbot understanding.
2Reliability
If the chatbot makes predictions about user intent, then dialogue generation becomes more accurate, but the system complexity increases due to neural networks and prediction models
Solution Approach 1:
The system implements self-service by using the chatbot's own prediction model to generate candidate questions based on user profile and conversation history. The model serves itself by automatically learning from conversation patterns and improving its predictions over time, eliminating the need for external manual programming of question generation logic.
Solution Approach 2:
The system incorporates feedback mechanisms where user selections of candidate questions, as well as subsequent conversation outcomes, are fed back into the prediction model. This feedback loop enables the neural network to continuously refine its intent prediction accuracy, improving dialogue generation reliability while managing complexity through data-driven optimization.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In certain embodiments, pre‑chat intent prediction and dialogue generation may be facilitated. In some embodiments, a chat initiation request may be obtained from a user. The latest activity information associated with the user may be provided to a prediction model to obtain predicted intents of the user. For each intent of the current intents, a candidate question may be selected from a question set based on the candidate question matching the intent. Within ten seconds of the chat initiation request, the candidate questions may be simultaneously presented on the chat interface. A user selection of a first question of the candidate questions may be obtained via the chat interface responsive to the presentation of the candidate questions. Based on the user selection and the first question matching a first intent of the predicted intents, the first intent may be provided as reference feedback for the prediction model.