Adaptive Auto-Reply System for Mobile Terminals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information replying methods on mobile terminals have a limited applicable scope and are inflexible in adapting to different language styles, leading to slow input speeds and inadequate user experience, especially in noisy environments or urgent situations.
Innovation Solution
An information auto-replying method and apparatus that utilize a database to search for dialog style information based on text content and contact information, perform preprocessing, and determine appropriate reply information, allowing for adaptive language style adjustment and user selection from candidate reply tuples, with the option to rerank and train models for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If speech recognition based input technology is used, then information input speed is greatly increased, but accuracy rate in noisy environment deteriorates and cannot reach practical level
Solution Approach 1:
The patent introduces an intermediary processing layer between speech recognition and final output. The system uses speech recognition to obtain initial text, then processes this text through a series of intermediate steps including extracting key information, matching with preset templates, and generating reply options. This intermediary processing resolves the contradiction by not relying solely on speech recognition accuracy but rather using it as one input among several to achieve accurate replies.
2Ease of operation
If preset rule or mode based input technology is used, then some user requirements are met, but applicable scope is limited and language style is fixed
Solution Approach 1:
The patent implements a dynamic reply generation system that adapts to different scenarios and communication partners. Instead of using fixed preset rules, the system dynamically selects and combines multiple processing methods including speech recognition results, text analysis, template matching, and user feedback. The reply generation adapts based on the communication context, partner relationship, and conversation history, thereby achieving both ease of operation and high adaptability.
Solution Approach 2:
The system changes multiple parameters to achieve adaptability: it adjusts the weight of different input sources (speech recognition confidence, template matching scores, user preferences), modifies the level of automation based on conversation context, and adapts the language style according to the communication partner and situation. These parameter changes enable the system to satisfy diverse user requirements while maintaining ease of operation.
3Measurement precision
If conventional manual information input is used, then language style can be precisely controlled, but information input speed is slow and user experience deteriorates
Solution Approach 1:
The patent applies preliminary action by pre-processing and storing multiple types of data including speech recognition results, text analysis outcomes, preset reply templates, and user communication patterns. Before actual reply generation is needed, the system has already prepared and organized this information, enabling fast retrieval and combination during the reply generation phase. This preliminary preparation resolves the contradiction by providing both speed (through pre-processing) and precision (through careful organization of language style options).
Data Source
AI summary
The information replying method includes: receiving to-be-replied information, where the to-be-replied information includes text content and contact information; searching a database for corresponding dialog style information according to the text content and the contact information; performing preprocessing on the text content, where the preprocessing includes word segmentation processing and stop word removal processing; and searching, according to data that has undergone the preprocessing, the database corresponding to the dialog style information, to determine reply information.