Generative AI Search Result Assembly to Reduce Database Queries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search systems face inefficiencies in processing search requests due to the overhead of contacting databases for search result data, which can be time-consuming and resource-intensive, and there is a need for improved mechanisms to handle search requests efficiently in terms of processing resources and response times.
Innovation Solution
A method utilizing a generative artificial intelligence model to generate first data fields and a machine-learning model to validate these fields, combined with a database system to determine second data fields, thereby optimizing search result assembly and reducing reliance on database queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional search systems contact databases for every search request to retrieve search result data, then data accuracy and completeness are maintained, but processing time increases and computational resources are consumed
Solution Approach 1:
The system performs preliminary actions by generating candidate search results using the generative AI model before database verification. This allows the system to prepare potential matches in advance, reducing the need for extensive database queries and thereby decreasing processing time while maintaining data accuracy through subsequent validation.
Solution Approach 2:
The generative AI model acts as an intermediary between the search request and the database system. It generates candidate search results that are then validated against the database, reducing the direct burden on the database system and minimizing query overhead while maintaining data accuracy through the validation step.
2Reliability
If traditional search systems contact databases for every search request, then data freshness is ensured, but computational load and resource consumption increase
Solution Approach 1:
The system applies partial action by validating only the generated candidate search results against the database rather than retrieving all possible data. This selective validation approach ensures data freshness for relevant results while significantly reducing computational load and resource consumption compared to comprehensive database queries.
Solution Approach 2:
The generative AI model creates copies or representations of potential search results before database verification. This allows the system to work with generated data locally rather than querying the database for every possible result, reducing computational load while maintaining data freshness through targeted validation.
3Productivity
If generative AI model is used to generate search result data, then processing speed improves, but data validity and accuracy may be compromised
Solution Approach 1:
The system implements feedback by validating the generated search results against the database system. The validation step provides feedback on the accuracy and validity of the AI-generated data, allowing the system to correct or reject invalid results while maintaining the processing speed benefits of generative AI for the majority of results.
4Reliability
If multiple data sources are used for search results, then data comprehensiveness improves, but system complexity increases
Solution Approach 1:
The system segments the search result generation process into two distinct parts: AI-generated candidate results and database-validated results. This segmentation allows the system to leverage multiple data sources comprehensively while managing complexity by clearly defining the role and workflow of each data source, with the AI model handling generation and the database handling validation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method, systems and computer programs for handling search requests at a search platform are provided. The search platform receives a search request determining a number of search results by generating one or more values for a first set of data fields by using a generative artificial intelligence model and determining, on the basis of the generated one or more values for the first set of data fields and by using the database system, one or more values for a second set of data fields. The search platform assembles a number of completed search results on the basis of the generated first set of data fields and the determined second set of data fields and returns at least one of the completed search results.