Question Answering System for Electronic Communication Response Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic communication tools hinder productivity due to the need for users to send and receive unnecessary messages that could be easily answered by a computer, leading to inefficiencies in real-time non-verbal communication.
Innovation Solution
A question answering system is integrated into electronic communication tools to generate candidate answers based on both personalized and informational data sources, allowing users to select pre-generated responses or block messages that can be easily answered, thereby reducing unnecessary communication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users manually type and send messages in electronic communication tools, then real-time non-verbal communication is enabled, but productivity is hindered due to unnecessary messages and time consumption
Solution Approach 1:
The system performs preliminary actions by generating candidate answers before the user needs to respond. The question answering system proactively queries data sources and prepares multiple potential responses in advance, allowing the user to simply select from pre-generated options rather than composing messages from scratch, thereby reducing time loss and improving productivity
Solution Approach 2:
The system enables self-service by automatically generating candidate answers without requiring manual composition. The question answering system autonomously queries personalized and informational data sources, processes the information, and presents ready-to-send responses, allowing users to communicate efficiently by selecting pre-prepared answers rather than manually typing each message
2Reliability
If a question answering system generates candidate answers from multiple data sources, then response accuracy and relevance are improved, but system complexity increases
Solution Approach 1:
The system applies segmentation by dividing the data sources into distinct categories: personalized data sources (user profiles, message histories, contacts) and informational data sources (encyclopedias, news, weather). This segmentation allows the system to manage complexity through modular data access while maintaining high answer accuracy by querying appropriate specialized sources for different types of questions
Solution Approach 2:
The system uses an intermediary approach by introducing a question answering system that acts as a mediator between the user and multiple data sources. This intermediary layer manages the complexity of querying multiple sources by providing a unified interface, processing queries through various data sources, and synthesizing results into candidate answers, thereby maintaining reliability without exposing the full system complexity to the user
3Measurement precision
If the system provides multiple selectable candidate answers to users, then communication precision and user control are improved, but the complexity of the communication interface increases
Solution Approach 1:
The system applies local quality by providing different levels of answer presentation based on the specific context and user needs. The interface displays candidate answers with varying degrees of detail and selection options appropriate to each situation, allowing users to exercise precise control when needed while maintaining simple operation for routine communications, thus balancing precision with ease of operation
Data Source
AI summary
Text is received from a first client. The text is associated with an electronic communication tool for communication to a second client. Candidate answers are generated based on the text using a question answering system. The question answering system generates the candidate answers based on a plurality of data sources, including at least one personalized data source and at least one informational data source. At least one of the candidate answers is provided to the second client. Each of the candidate answers provided to the second client is selectable.