Collaborative File Relevance via Human Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In collaborative file storage systems, existing search engines often produce lists of related documents that include irrelevant results, as they rely on algorithms that fail to recognize the relevance of newer documents and may miss valuable recent updates, leading to a lack of accuracy in suggesting relevant files to users.
Innovation Solution
A method where a system server identifies files related to a first file by generating inquiries for collaborators, who provide responses to select and rank relevant files, creating hyperlinks between the first file and selected files, and storing this information for future searches, thereby improving the relevance of search results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If search engines use algorithms to generate related documents, then the system can automatically provide search results, but the relevance and accuracy of the results deteriorate due to inclusion of irrelevant documents and failure to recognize newer updates
Solution Approach 1:
The system implements feedback by sending inquiries to collaborators about the relevance of search results and using their responses to refine and re-rank the document list, thereby improving accuracy while maintaining automated operation
Solution Approach 2:
Collaborators serve as intermediaries between the search engine algorithm and the final search results, providing human judgment to filter and rank documents, thus resolving the contradiction between automated generation and relevance accuracy
2Quantity of substance
If the system generates a large number of search results, then more potential relevant documents are included, but the difficulty of sorting and identifying relevant documents increases
Solution Approach 1:
Collaborators provide feedback on the relevance of each document in the search results, enabling the system to re-rank documents based on human judgment rather than just algorithmic scoring, making it easier for users to identify relevant documents
Solution Approach 2:
The system changes the ranking parameter from algorithmic relevance scores to collaborator-assessed relevance, transforming the ordering criterion to improve ease of identifying relevant documents while maintaining a comprehensive result set
3Measurement precision
If authors suggest related documents manually, then the relevance quality improves, but the documents suggested are older than the selected document and cannot provide recent updates
Solution Approach 1:
The system performs preliminary automated identification of potentially relevant documents including recent updates, then uses collaborator feedback to filter and confirm relevance, combining the speed of automated detection with the quality of human judgment
Solution Approach 2:
Collaborators act as intermediaries to validate the timeliness and relevance of recently published documents, allowing the system to include current updates while maintaining high relevance quality through human assessment
Data Source
AI summary
Methods and systems for providing related files in a collaborative file storage system are disclosed. One method includes identifying a plurality of files within the collaborative file storage system, wherein the plurality of files each have a relationship with the first file, and wherein the collaborative file storage system allows sharing of the plurality of files between multiple users through a network. The method further includes generating, by a system server, a list of inquiries based on the plurality of files, providing, by the system server, the list of inquiries to at least one collaborator of the first file, receiving from the at least one collaborator at least one response to the list of inquiries, selecting a subset of the plurality of files based on the at least one response, and storing information related to the selected subset of the plurality of files for access if the first file is selected.