Neural Network Query Parser for Conceptual Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional database systems lack the ability to process conceptual queries effectively, as they fail to consider the context associated with search queries not input in a recognized query format, leading to inefficient searches and irrelevant results.
Innovation Solution
Implementing an artificial intelligence model, such as a neural network, that encodes search queries on a per-word and per-character level, uses Named Entity Recognition to tag query components, and transforms conceptual queries into standardized database queries, enabling context-based filtering and retrieval of specific data objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional database systems use keyword searches or database queries in recognized formats, then data retrieval can be performed, but the system cannot process queries with data values or information not in recognized query formats, leading to limited capabilities
Solution Approach 1:
The patent introduces an intermediary component (query parser/interpreter) that sits between the user input and the database execution engine. This intermediary translates natural language conceptual queries into structured query formats that the database can execute, enabling the system to handle queries in unrecognized formats while maintaining the existing database execution architecture.
Solution Approach 2:
The system dynamically changes the parameters of query processing by accepting variable input formats (natural language, semi-structured text) and transforming them into standardized query parameters. This allows the database system to adapt to different query styles without requiring complete architectural redesign.
2Measurement precision
If the system performs keyword searches without considering context, then simple queries can be executed quickly, but the results are irrelevant when the query is not in a recognized format and context is needed
Solution Approach 1:
The system performs preliminary actions by pre-parsing and pre-processing the query input before executing the full database search. This includes identifying key entities, relationships, and constraints in the query, and preparing a structured representation that guides the subsequent search process, thereby improving result relevance without excessive time loss.
Solution Approach 2:
The query processing is segmented into multiple stages: initial parsing to identify query components, context analysis to understand relationships between components, and execution to retrieve results. This segmentation allows the system to apply different levels of processing intensity to different parts of the query, improving overall precision while managing time consumption.
3Loss of information
If the system only processes queries in recognized formats, then processing is straightforward, but it fails to identify user intent and context in natural language queries
Solution Approach 1:
An intermediary natural language processing layer is introduced that acts as a mediator between the simple keyword search mechanism and the complex task of understanding user intent. This layer extracts semantic meaning, entities, and relationships from natural language queries without requiring complete rewriting of the underlying search infrastructure.
Solution Approach 2:
The system adds another dimension to query processing by incorporating semantic analysis and context understanding layers above the traditional keyword matching dimension. This allows queries to be processed both as simple keyword searches and as meaningful natural language statements, capturing user intent without abandoning the efficiency of traditional search methods.
Data Source
AI summary
Methods, systems, and devices supporting detecting and processing conceptual queries are described. A device (e.g., an application server) may receive a search query from a user device. The search query may include one or more parameters. The device may tag the search query using one or more tags associated with the one or more parameters. In some examples, the one or more tags may be determined based on a neural network. The device may determine that the search query is supported as a conceptual query based on a tag of the one or more tags corresponding to a data object stored in a database. The device may then generate a database query in a query language based on the search query, retrieve a set of results for the search query using the database query in the query language, and transmit the set of results to the user device.