Semantic Application Search Using Feature Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users struggle to find specific applications on electronic devices due to forgetting their names and the inefficiency of current keyword-based search methods, leading to time-consuming searches through numerous applications.
Innovation Solution
An application search method utilizing semantic feature vectors, including categories and expanded related words, generated by a natural language model, to determine semantic similarities between user queries and application information, enabling more accurate and efficient search results without requiring exact application names.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If users search for applications by entering keywords, then they can find applications, but they need to input keywords very precisely and spend time trying multiple keywords
Solution Approach 1:
The patent introduces semantic feature vectors as an intermediary between user queries and application names. Instead of requiring direct keyword matching, the system converts both queries and applications into semantic feature vectors, then compares these vectors to find matches. This intermediary layer enables semantic search that understands meaning rather than requiring exact keyword matches.
Solution Approach 2:
The patent changes the search parameter from exact keyword matching to semantic similarity comparison. By transforming application names and user queries into semantic feature vectors, the system can compare them based on semantic similarity rather than exact string matching, allowing for more flexible and accurate search results.
2Reliability
If users sift through a vast list of applications one by one, then they can find the desired application, but the search process becomes time-consuming
Solution Approach 1:
The patent replaces the mechanical manual browsing process with an automated semantic search system. Instead of users manually sifting through application lists, the system automatically computes semantic feature vectors, compares them with query vectors, and returns relevant results, substituting manual inspection with computational semantic analysis.
Solution Approach 2:
The system performs preliminary action by pre-computing and storing semantic feature vectors for all applications before user queries arrive. This allows the system to quickly compare new queries against pre-prepared data without requiring real-time analysis of each application during the search process.
3Ease of operation
If users forget application names, then they cannot activate the required application immediately, but they can still search by entering keywords
Solution Approach 1:
The patent uses semantic feature vectors as an intermediary that bridges the gap between user memory limitations and application identification. Even when users forget exact application names, they can enter partial or vague keywords, and the semantic feature vector comparison system can still accurately identify the intended application based on semantic meaning rather than exact naming.
Data Source
AI summary
An electronic device and an application search method thereof are provided. The method includes the following steps. A plurality of semantic feature vectors of each of a plurality of application is recorded. The semantic feature vectors of each of the applications at least include a first semantic feature vector associated with one of application categories and a second semantic feature vector associated with one of expanded related words. A query string is obtained via the input device. A query semantic feature vector of the query string is generated by using a natural language model. A plurality of semantic similarities between the query semantic feature vector and the semantic feature vectors of each application are determined. A target application corresponding to the query string is determined from the applications according to the semantic similarities corresponding to the semantic feature vectors of each application.