Evaluation Data Query System for Entity Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data query methods fail to provide effective search results for users seeking evaluations of specific products with certain characteristics due to the proliferation of low-quality and scattered data on the internet, often contaminated by 'water armies' and trolls, leading to invalid evaluations.
Innovation Solution
A method and device for data query based on evaluation that aggregates and filters data from multiple websites, extracting labels indicating user views, performing quality scoring and sensitive word marking, and aggregating evaluation data by type and label to provide comprehensive and accurate information on target entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If evaluation data is collected from multiple websites, then the quantity of evaluation data increases, but the quality of evaluation data decreases due to water armies and trolls
Solution Approach 1:
The patent segments the evaluation data processing into multiple stages: data collection from multiple websites, quality filtering to remove water army and troll evaluations, label extraction to categorize genuine user views, and aggregation to synthesize reliable evaluation information. This segmentation allows the system to handle large quantities of data while maintaining quality through systematic processing at each stage.
Solution Approach 2:
The patent introduces an intermediary filtering mechanism that acts as a mediator between raw evaluation data and final results. The filter evaluates each piece of data for authenticity, removing low-quality evaluations from water armies and trolls while preserving genuine user feedback. This intermediary layer resolves the contradiction by selectively processing data based on quality criteria.
2Ease of operation
If basic search query is used, then the operation simplicity is maintained, but the search accuracy decreases for specific product characteristics
Solution Approach 1:
The patent performs preliminary actions by pre-extracting labels from evaluation data and organizing them by product type and characteristics before users perform searches. This pre-processing creates a structured database of labeled evaluation information, allowing users to perform simple queries while the system automatically retrieves and presents relevant, accurate information matching specific product characteristics.
Solution Approach 2:
The patent replaces the mechanical search process with an intelligent information retrieval system. Instead of requiring users to manually filter and analyze evaluation data, the system automatically extracts labels, filters quality, and aggregates information based on user queries. This substitution maintains operational simplicity while dramatically improving search accuracy through automated label-based retrieval.
3Loss of information
If all evaluation data is processed, then the completeness of information is improved, but the processing time increases due to data volume
Solution Approach 1:
The patent extracts only the essential and relevant information from evaluation data by identifying and extracting labels that represent key user views and product characteristics. Instead of processing and storing all raw evaluation data, the system extracts meaningful labels and aggregates them, significantly reducing processing time while maintaining information completeness for search purposes.
Solution Approach 2:
The patent applies partial action by processing only the necessary portions of evaluation data required to answer user queries. The label extraction and aggregation process focuses on retrieving specific information related to product characteristics mentioned in queries, rather than uniformly processing all available data. This selective processing maintains information completeness for relevant queries while minimizing unnecessary processing time.
Data Source
AI summary
A method of data query based on an evaluation and a device. The method includes: obtaining evaluation data of entities and basic information of the entities from multiple websites; extracting labels of the entities according to the evaluation data of the entities and the basic information of the entities; filtering the evaluation data of the entities; aggregating to obtain evaluation data of the same type of entities having the same label according to the basic information of the entities, the labels of the entities and the filtered evaluation data of the entities; and making a query to obtain information on a target entity according to a retrieval statement and the aggregated evaluation data of the same type of entities for each label. The method can make a query for information on a type of entities for a user, so that the user can compare the entities.