Automated Software Query Generation for Search Engine Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack an automated method to test software component search engines, resulting in significant manual effort and errors in validating their accuracy, making the process time-consuming and prone to inaccuracies.
Innovation Solution
A system and method for automatically generating search queries from software documents to validate software component search engines, utilizing a processor-driven sequence of instructions that includes a Web GUI portal, Description Readme and Code Parser, Tokenize and Preprocessing Engine, Query Generator, Summarizer Service, Splitter and Resizer Service, Paraphraser Service, and Sentence Corpus Machine Learning Service to parse, process, and generate search queries, thereby automating the validation process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual methods are used to test software component search engines, then validation can be performed, but significant manual effort and time are required, and errors in interpreting accuracy occur
Solution Approach 1:
The system enables self-service automation by having the search engine validate itself through automated query generation and execution. The validation system automatically generates search queries, executes them against the search engine, and evaluates results without requiring manual intervention, thus resolving the contradiction between automation extent and time loss
Solution Approach 2:
The system performs preliminary actions by pre-generating a large set of search queries and organizing them into test suites before actual validation execution. This preliminary preparation enables rapid automated validation, reducing the time required during actual testing while maintaining high automation levels
2Measurement precision
If manual methods are used to validate search engine accuracy, then validation can be performed, but significant manual errors occur in interpreting accuracy
Solution Approach 1:
The system implements automated feedback mechanisms where the validation system automatically evaluates search engine results, compares them against expected outcomes, and generates precision metrics. This automated feedback loop eliminates manual interpretation errors while maintaining high measurement precision in accuracy validation
Solution Approach 2:
The system replaces manual mechanical interpretation processes with automated computational analysis. Instead of human reviewers manually evaluating and interpreting search results, the system uses automated algorithms to measure and interpret accuracy, eliminating human errors while improving measurement precision
3Reliability
If a comprehensive test suite is created manually, then coverage can be achieved, but the process is extremely time consuming
Solution Approach 1:
The system performs preliminary action by automatically generating and organizing comprehensive test queries into structured test suites before validation execution. This pre-preparation enables complete coverage to be achieved rapidly during actual validation, resolving the contradiction between reliability and productivity
Solution Approach 2:
The system maintains continuity of useful action by automatically and continuously generating validation queries and executing tests without manual interruption. This continuous automated process achieves comprehensive coverage while maintaining high productivity, as the system can execute large volumes of tests without the breaks and delays inherent in manual processes
Data Source
AI summary
Systems and methods for automatically detecting search queries from software documents to validate component searches are provided. An example method includes receiving a number of documents, parsing a number of documents to extract a number of sentences, mapping each of the number of sentences to a number of sections, filtering the number of sentences, consolidating the number of sections and the number of filtered sentences into a summary, partitioning the summary into a first number of sentences, replacing one or more of the number of sentences with one or more paraphrase sentences generated by a machine learning model to generate a second number of sentences, and generating a number of search queries based on the second number of sentences.