Word Completion Trie Structure for Search Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional word completion methods face inefficiencies when users enter short character strings, leading to excessive hot word suggestions and poor user experience due to low search efficiency and unclear completion triggers.
Innovation Solution
An improved trie structure is implemented, where hot words are stored in nodes based on their completion probability, and only words with high completion probabilities are retained, allowing for efficient search and recommendation of relevant hot words by matching character strings entered by users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If prefix matching is performed on all hot words based on user input, then comprehensive word suggestions are provided, but search efficiency deteriorates and excessive suggestions are generated when input is short
Solution Approach 1:
The patent pre-calculates and stores completion probabilities for all hot words in the trie structure during system initialization. This preliminary action allows the system to quickly retrieve pre-computed probability values during user input processing, avoiding real-time calculation overhead and improving search efficiency while maintaining suggestion quality
Solution Approach 2:
The patent applies different completion probability thresholds to different nodes in the trie structure based on their local characteristics. Nodes representing shorter input prefixes use higher thresholds to filter out low-quality suggestions, while nodes for longer inputs use lower thresholds. This localized quality control optimizes both efficiency and suggestion relevance for different input lengths
2Reliability
If all hot words meeting prefix matching are returned, then comprehensive suggestions are provided, but user experience deteriorates due to excessive and irrelevant suggestions
Solution Approach 1:
The patent dynamically adjusts the completion probability threshold parameter based on the length of user input. When users enter short character strings, a higher threshold is applied to filter out low-probability suggestions. As input length increases, the threshold decreases to provide more suggestions. This parameter adaptation resolves the contradiction between coverage and user experience by providing appropriately filtered results for different input scenarios
Solution Approach 2:
The patent creates a simplified copy of the hot word database in trie structure format, organized by prefixes. This copied structure enables efficient navigation and filtering without processing the entire original database, allowing quick retrieval of only relevant high-probability suggestions that match user input, thereby improving user experience through faster, more targeted results
3Reliability
If hot words with low completion probability are included, then comprehensive suggestions are provided, but completion efficiency deteriorates due to processing of low-value words
Solution Approach 1:
The system performs preliminary filtering during trie construction by calculating completion probabilities for all potential hot words and storing only those meeting minimum thresholds in each node. This pre-filtering action eliminates low-value words before they can consume processing time during user queries, while still maintaining a comprehensive database of all valid hot words in the underlying storage
Solution Approach 2:
The patent extracts and removes hot words with completion probabilities below the threshold from the searchable trie nodes. By taking out these low-value words from the active suggestion pool while preserving them in the underlying database, the system reduces processing time and memory usage during completion operations without permanently losing access to the complete hot word database
Data Source
AI summary
A word completion method and apparatus are described. The method is applied to a search scenario and is used to complete an incomplete word entered by a user. The method is based on an improved trie. Hot words are stored in some nodes of the trie. In the word completion method, a target node that matches a character string is searched in the trie, and at least one completed word is output to a user based on a hot word stored in the target node. Word completion efficiency is improved, and a case in which a word is recommended to a user when the user enters an excessively short character string is avoided.