Morphological Query Expansion for Database Search Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current database search systems face inefficiencies when handling search strings with multiple possible alternatives due to misspellings, variant transliterations, or abbreviations, leading to incomplete results and prolonged processing times, especially in name searches where accuracy and speed are critical.
Innovation Solution
A database searching method that identifies all possible matches for a search string by using a unique identifier system and morphological rules to account for multiple morphologies, ranking results based on probability, and comparing elements against a normalized database index to ensure comprehensive and efficient retrieval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional database search methods are used to handle search strings with multiple possible alternatives, then the search can be performed, but the processing time is prolonged and results may be incomplete
Solution Approach 1:
The patent applies preliminary action by pre-processing search strings to generate all possible alternative forms (misspellings, variant transliterations, abbreviations) before the actual database search. This includes creating a set of alternative search terms and expanding the search query to include multiple morphologies upfront, so that all possible matches can be retrieved in a single search operation rather than requiring multiple iterative searches.
Solution Approach 2:
The patent segments the search string into multiple alternative forms and processes each segment independently. By dividing the search query into base terms and their alternatives, the system can efficiently search for each segment and combine results, reducing the overall processing time while maintaining comprehensive coverage of all possible matches.
2Measurement precision
If comprehensive search is performed to find all possible matches including variant spellings, then accuracy improves, but processing complexity increases
Solution Approach 1:
The patent changes parameters by transforming the search string into multiple alternative forms with different morphological parameters. This includes generating alternatives with different spellings, transliterations, and abbreviations, then using these transformed parameters to query the database. The system manages complexity by systematically varying these parameters rather than using complex algorithms to handle each case individually.
Solution Approach 2:
The patent introduces an intermediary processing layer that sits between the user's search query and the database search execution. This intermediary generates and manages the set of alternative search terms, handling the complexity of variant spellings and transliterations centrally. The intermediary prepares the expanded search query with all alternatives before passing it to the database engine, simplifying the overall system architecture.
3Measurement precision
If morphological query term expansion is used to handle all search string alternatives, then search completeness improves, but processing time increases
Solution Approach 1:
The patent performs morphological expansion in advance by generating all alternative forms of search terms before the actual database search. This preliminary action includes creating abbreviations, correcting misspellings, and generating variant transliterations upfront, so that the database search can be executed once with a comprehensive query rather than requiring multiple sequential searches for each alternative.
Solution Approach 2:
The patent merges multiple search queries into a single comprehensive search operation. By combining all alternative search terms and their variations into one unified query, the system retrieves all possible matches in a single database pass, improving search speed while maintaining completeness. The results are then consolidated and ranked to provide the final comprehensive answer.
Data Source
AI summary
The present invention is a method of database searching. First, a language is selected. Next, elements are received. The system is then searched to identify at least one unit number that is associated with a chosen element, the unit number being linked to a data unit containing morphological variants of the element. If no unit number is identified, the element is compared to a prefix list. If no match is found there, the element is broken into a prefix and suffix, and the prefix and suffix are matched to a prefix list, suffix list or a unit number. This process is repeated for all elements. A unit number associated with each element is then chosen, and the elements contained in the data units linked to the unit numbers are compared to a database. The results are displayed and preferably ranked according to user preferences. If an element is associated with multiple unit numbers, this process is repeated until all data units have been compared to the database.