Constraint-Aware Search Ranking with Incremental DAG Preferences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing database management systems and information retrieval systems lack the ability to express mandatory search constraints in natural language and do not efficiently handle progressive discovery of search results from multiple databases, leading to inefficiency and impracticality in large-scale industrial applications.
Innovation Solution
A method and system that allows for the progressive computation of search results across multiple databases, utilizing artificial intelligence techniques, combinatorial optimization, and natural language processing to represent preferences as directed acyclic graphs, enabling efficient and scalable retrieval of optimal and sub-optimal solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If centralized systems are used to compute optimal solutions with preferences, then absolute and global ranking of search results is achieved, but computation must restart from the beginning when new results arrive, leading to significant inefficiency
Solution Approach 1:
The patent segments the computation process into independent components: (1) computing optimality scores for newly arrived search results using the incremental method, and (2) merging these new results with the existing ranked list. This segmentation allows the system to avoid restarting the entire computation from scratch, thereby maintaining measurement precision while significantly improving computation efficiency.
Solution Approach 2:
The patent performs preliminary actions by pre-computing and storing the optimality score function and the existing ranked search results before new results arrive. When new search results are obtained, the system can immediately compute their optimality scores using the pre-established function and insert them into the appropriate positions in the ranked list, rather than re-computing the entire ranking from scratch.
2Ease of operation
If methods calculate scores to associate with search results for relative ranking, then ranking of results is achieved, but new search results dynamically emerging require restarting computation from the beginning
Solution Approach 1:
The patent implements a dynamic ranking system where the optimality score function is designed to be incremental. When new search results emerge, the system dynamically updates the ranked list by computing scores only for the new results and inserting them at appropriate positions, rather than static re-computation of all results. This dynamic approach maintains ease of operation while reducing computation time.
Solution Approach 2:
The patent changes the parameter representation by using an optimality score function that can evaluate new results independently. Instead of maintaining complex inter-dependent rankings, the system uses scoring parameters that can be computed incrementally for new results and merged with existing rankings, thereby reducing the time loss when new search results arrive.
3Reliability
If automated reasoning with preferences based on dominance tests is used, then optimal solutions are identified, but all admissible solutions must be available at the start of the process
Solution Approach 1:
The patent segments the solution identification process into two independent phases: (1) computing optimality scores for newly arrived admissible solutions using the incremental method, and (2) updating the set of optimal solutions by comparing new scores with existing ones. This segmentation allows the system to maintain reliability in identifying optimal solutions while adapting to progressively arriving results without requiring all solutions to be available upfront.
Solution Approach 2:
The patent introduces an optimality score function as an intermediary between new search results and the set of optimal solutions. This intermediary function enables the system to evaluate and integrate new results incrementally, bridging the gap between dominance-based optimal solution identification and the ability to handle progressively arriving results from multiple databases.
4Quantity of substance
If systems query multiple databases with progressive result availability, then comprehensive search results are obtained, but existing methods require restarting computation on extended sets
Solution Approach 1:
The patent segments the computation process into independent incremental steps that can handle expanding result sets. Instead of recomputing over the entire extended set when new results arrive from additional databases, the system segments the work into computing scores for new results and merging them with existing rankings, thereby maintaining computation efficiency while increasing the quantity of search results.
Solution Approach 2:
The patent ensures continuity of useful action by maintaining the optimality score function and ranked list structure throughout the process of incorporating results from multiple databases. This continuous structure allows new results to be integrated incrementally without interrupting or restarting the computation, thereby improving productivity while handling increasing quantities of search results from multiple sources.
Data Source
AI summary
In one aspect, a method for determining personalized search results with respect to mandatory constraints and qualitative preferences on search criteria representable via a DAG includes collecting input from users or software systems; transforming the inputs into the mandatory constraints and preferences on search criteria, wherein the search criteria is partially ordered and representable through the DAG; determining weights for the search criteria using a function, and associating the weights with the nodes of the DAG to determine a partial ordering of such weights equivalent to an order of the DAG nodes, where any node at level “k” has a weight “w” such that sum of all or any finite subset of the weights of the nodes at levels below “k” is always less than “w”, the level being defined by a topological order of the DAG; and determining search results that satisfy the mandatory constraints on the search criteria.