Search Suggestion Scoring via Entity Co-occurrence Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current search engines provide inaccurate search suggestions due to historic keyword references that can refer to multiple topics, leading to irrelevant results when users search for entities like organizations or locations, resulting in a need for quicker and more accurate suggestions.
Innovation Solution
A method employing entity and feature co-occurrence analysis using an entity database and a trends database to generate search suggestions, where entities are identified, scored, and combined to provide a ranked list of relevant results, with on-the-fly learning to improve search accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If search engines use historic keyword references to generate search suggestions, then the system can provide forecast features like query auto-complete, but the suggestions become inaccurate because one keyword can refer to multiple topics
Solution Approach 1:
The patent segments the search suggestion generation process into multiple independent components: entity extraction module, feature extraction module, co-occurrence analysis module, and scoring module. Each module handles a specific aspect of the problem, allowing the system to disambiguate keywords by analyzing them in context rather than treating them as single-topic indicators.
Solution Approach 2:
The patent introduces co-occurrence data as an intermediary element that connects keywords to their actual contextual meanings. By analyzing which entities and features co-occur with search keywords in the training corpus, the system can determine the most likely intended topic even when a keyword has multiple possible meanings.
2Quantity of substance
If search engines return results based on mixed entities with the same or similar names, then the system can retrieve a large volume of documents, but the results become irrelevant to the user's actual interest
Solution Approach 1:
The patent applies local quality by assigning different relevance scores to different entities based on their specific contextual characteristics. Instead of treating all entities uniformly, the system analyzes the local context (co-occurring features and entities) to determine which entity matches the user's intent, thereby filtering out irrelevant results while maintaining comprehensive coverage.
Solution Approach 2:
The patent changes the parameter of entity selection by introducing a scoring mechanism that dynamically adjusts entity relevance based on co-occurrence frequency and feature matching. This allows the system to prioritize the most relevant entities while still considering a broad range of possibilities, balancing quantity and quality of results.
3Measurement precision
If search engines use traditional keyword-based suggestion methods, then the implementation is simple, but the suggestions are not quicker or more accurate
Solution Approach 1:
The patent applies preliminary action by pre-processing the training corpus to extract entities, features, and co-occurrence relationships before the actual search suggestion generation. This pre-computation creates structured data that can be quickly queried during search operations, improving both accuracy and speed while making the complex analysis manageable through staged processing.
Data Source
AI summary
A method for obtaining and providing search suggestions using entity co-occurrence is disclosed. The method may be employed in any search system that may include at least one search engine, one or more databases including entity co-occurrence knowledge and trends co-occurrence knowledge. The method may extract and disambiguate entities from search queries by using an entity and trends co-occurrence knowledge in one or more database. Subsequently, a list of search suggestion may be provided by each database, then by comparing the score of each search suggestion, a new list of suggestion may be built based on the individual and/or overall score of each search suggestion. Based on the user's selection of the suggestions, the trends co-occurrence knowledgebase can be updated, providing a means of on-the-fly learning, which improves the search relevancy and accuracy.