Database Search Relevance via Interaction Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database search methods often return irrelevant data sets due to ambiguous search terms or method artifacts, causing relevant results to be buried deep in the list, leading to user frustration and loss for web shops and other applications.
Innovation Solution
A method that captures user interactions with search results, generates correlation data between attribute interactions, and iteratively updates relevance data to improve search query accuracy, using machine learning and AI to enhance relevance determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional search methods are used to search databases, then the search process is simple and fast, but irrelevant data sets are returned and relevant hits are buried deep in the list
Solution Approach 1:
The patent implements feedback by capturing user interactions with search results and using this information to iteratively update correlation data and relevance scores. The system monitors how users interact with displayed data sets and feeds this information back into the search algorithm to improve future relevance determinations, thereby resolving the contradiction between accuracy and complexity by making the system adaptive rather than statically complex
Solution Approach 2:
The search system performs self-service by automatically learning from user interactions without requiring manual intervention or reconfiguration. The system autonomously updates correlation data between attributes and interaction patterns, automatically improving its relevance determination capabilities over time, thus achieving high accuracy without proportionally increasing operational complexity
2Measurement precision
If iterative learning from user interactions is implemented, then relevance determination accuracy is improved, but computational complexity and data processing requirements increase
Solution Approach 1:
The patent applies partial action by focusing computational resources on processing only the interaction data that is most relevant to improving search accuracy, rather than analyzing all possible data. The system selectively updates correlation data based on meaningful user interactions, achieving improved relevance determination without requiring excessive computational power to process every possible data point
Solution Approach 2:
The system performs preliminary action by pre-computing and storing correlation data between attributes and interaction patterns during periods of lower demand, then using this pre-processed information during actual search operations. This approach reduces the computational burden during real-time search while still achieving accurate relevance determination through the pre-analyzed correlation data
3Reliability
If correlation data from user interactions are incorporated into relevance data, then search result relevance is improved, but the system requires more sophisticated data processing capabilities
Solution Approach 1:
The patent merges correlation data derived from user interactions with traditional relevance data in an integrated manner. Rather than maintaining separate complex processing systems for different data types, the system combines attribute correlation data, interaction data, and traditional relevance factors into a unified relevance determination framework, achieving improved reliability while managing complexity through integration rather than separation
Data Source
AI summary
A method for searching a database having data sets (Mi) which comprise attributes (A1i to Ani). If a search query (S) is captured, the data sets (Mi) are assigned relevance data (RSi) for the search query (S) and at least one subset of the data sets (Mi) is output on the basis of the relevance data (RSi). Relevance data (RSi) of the data sets (Mi) for the search query (S) are generated with the inclusion of correlation data (KS,1 to KS,n) and the attributes (A1i to Ani) of the data sets (Mi). A computer program having program code for carrying out the method according to the invention when the program code is executed by a computer, a search engine, and a system having a search engine.