Query Classifier for Intrinsically Diverse Session Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional search engines lack support for intrinsically diverse information retrieval intents, requiring users to issue multiple queries to find different aspects of a topic, leading to inefficient information retrieval.
Innovation Solution
A system that predicts and optimizes search results for intrinsically diverse sessions by employing a query classifier to identify and classify queries, re-ranking results, and presenting supplemental user interface elements to provide relevant information across multiple facets of a topic with fewer queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional search engines provide support for extrinsic diversity to cover diverse information retrieval intents of many users, then the search results can cover multiple topics (e.g., Cardinals baseball, Cardinals football, Catholic Cardinals), but the system cannot effectively support intrinsically diverse sessions directed towards a single user's complex task spanning multiple aspects
Solution Approach 1:
The patent segments intrinsically diverse sessions into multiple subtasks, where each subtask represents a specific aspect of the user's complex information retrieval intent. The query classifier divides the session into subtasks and identifies aspect queries, allowing the system to handle each aspect separately while maintaining overall session coherence. This segmentation resolves the contradiction by providing structured support for diverse intents without overwhelming system complexity.
Solution Approach 2:
The patent employs preliminary action by using a query classifier to predict whether a query belongs to an intrinsically diverse session and to identify its aspect before retrieving search results. This preliminary classification enables the system to prepare appropriate retrieval strategies in advance, supporting diverse information retrieval intents efficiently without ad-hoc complexity management.
2Productivity
If conventional approaches optimize and evaluate single query sessions, then the system can handle individual queries effectively, but users must issue multiple queries to obtain documents on different aspects of a question or topic
Solution Approach 1:
The patent merges multiple aspect queries into a unified intrinsically diverse session model. By identifying that multiple queries belong to the same session and categorizing them as aspects of a single complex task, the system can retrieve and present documents covering all aspects in one coordinated operation rather than requiring separate sequential queries, thus reducing the number of queries needed and time loss.
Solution Approach 2:
The system uses feedback from the query classifier to adapt its retrieval strategy. When the classifier identifies a query as part of an intrinsically diverse session and determines its aspect, the system adjusts its document retrieval and presentation to ensure comprehensive coverage of all aspects, improving productivity by reducing the need for additional follow-up queries.
3Loss of time
If the system identifies and optimizes search results for intrinsically diverse sessions, then the number of queries needed to find diverse information is reduced, but the system must evaluate objective functions and compute optimized values to present results according to the optimized value
Solution Approach 1:
The patent changes the parameter space by introducing an objective function that evaluates multiple aspects of query sessions (e.g., aspect coverage, result diversity). By computing optimized values based on these parameters, the system can systematically reduce the number of queries needed while managing computation complexity through structured parameter evaluation rather than unstructured processing.
Solution Approach 2:
The objective function serves as an intermediary between the query classifier and the search result presentation. It translates the complex requirements of intrinsically diverse session optimization into computable metrics and optimized values, mediating the computation process to reduce time loss without directly exposing the full complexity of the optimization problem.
Data Source
AI summary
Various technologies described herein pertain to predicting intrinsically diverse sessions and retrieving information for such intrinsically diverse sessions. Search results retrieved by a search engine responsive to executing a query are received. A query classifier can be employed to determine whether the query is intrinsically diverse or not intrinsically diverse based on one or more features of the query and session interaction properties. The query is intrinsically diverse when included in an intrinsically diverse session directed towards a task, where the query and disparate queries included in the intrinsically diverse session are directed towards respective subtasks of the task. An objective function can be evaluated based at least upon the query to compute an optimized value when the query is determined to be intrinsically diverse. The search results can be presented on a display screen according to the optimized value when the query is determined to be intrinsically diverse.