Search Result Removal Using Attribute-Based Batch Exclusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing search technologies fail to adequately exclude or remove particular search results for a given query, leading to user experience issues, inaccurate search results, and excessive computing resource consumption.
Innovation Solution
Implement a system that automatically marks and removes search results based on user input via a UI element, utilizing machine learning models to determine item listing attributes, categories, and similarities, reducing the need for manual selection and repetitive queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing search technologies use algorithms to rank and retrieve search results, then search results can be retrieved, but the system fails to adequately exclude or remove particular search results and consumes excessive computing resources
Solution Approach 1:
The system performs preliminary actions by automatically marking search results as candidates for removal before the user needs to manually review each result. The machine learning model analyzes search results in advance and pre-identifies which results should be excluded, so when the user selects the removal option, only the pre-identified candidates are processed, significantly reducing computing resource consumption while maintaining accurate search result exclusion.
Solution Approach 2:
The search system provides self-service by automatically detecting and marking search results that should be removed based on user feedback patterns and machine learning analysis. Instead of requiring users to manually identify and remove unwanted results, the system autonomously determines which results are candidates for removal, reducing the computational burden on both the user and the system while improving search result accuracy.
2Reliability
If users manually select and remove search results one by one, then search result accuracy can be improved, but user experience deteriorates due to time consumption and repetitive queries
Solution Approach 1:
The system merges multiple individual search result removal operations into a single batch processing operation. When a user selects the removal option, the machine learning model identifies all search results that should be removed simultaneously based on their characteristics and user feedback patterns, allowing multiple results to be removed in one action rather than requiring separate manual selections for each result, thereby significantly reducing time loss while maintaining accuracy.
Solution Approach 2:
The system implements feedback mechanisms where user interactions with search results (such as viewing, selecting, or rejecting specific results) are fed back into the machine learning model. This feedback loop enables the system to learn which search results are most likely to be unwanted and automatically mark them for removal, improving search result accuracy over time while reducing the manual effort and time required from the user.
3Measurement precision
If the system processes each search result individually for removal, then removal accuracy can be maintained, but computing resource consumption and network latency increase
Solution Approach 1:
The system performs preliminary analysis of search results using machine learning models before the actual removal process. The model pre-identifies which search results are candidates for removal based on their attributes and user feedback patterns, so that when removal is executed, only the pre-identified candidates are processed individually. This preliminary action maintains removal accuracy for the subset of results that need to be removed while significantly reducing overall computing resource consumption and network latency compared to processing all results individually.
Data Source
AI summary
Various embodiments improve search technologies and computer information retrieval by automatically marking at least a first item listing, of a first set of item listings, as a candidate for removal as a search result for a query. Such automatic marking occurs in response to receiving an indication that a selection has been made at a computing device, where the selection is at least partially indicative of the user requesting removal, from a set of search results, of a first item listing based on a particular attribute value associated with the first item listing.