Media Post Tag Selection Using Context-Aware Candidate Tags
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for manually inputting tags to media content are inefficient and result in suboptimal tag quality, affecting the efficiency and quality of content posting.
Innovation Solution
A method and system that displays a set of candidate tags in a post interface, determined based on input information, allowing users to select a target tag, which is then displayed in the review page of the posted media content, enhancing tag accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users manually input tags for media content, then users have flexibility in tag selection, but the process is inefficient and tag quality is suboptimal
Solution Approach 1:
The system performs preliminary analysis of the media content (image, video, or audio) to automatically generate a set of candidate tags before the user needs to select tags. This pre-computation of relevant tags based on content analysis resolves the contradiction by providing ready-to-use tag options that improve both efficiency (users don't need to manually think of tags) and accuracy (tags are derived from actual content features)
2Measurement precision
If the system provides a list of candidate tags, then tag accuracy improves, but the interface complexity increases
Solution Approach 1:
The tag selection process is segmented into distinct stages: (1) system automatically generates candidate tags from content analysis, (2) system presents a curated list of candidate tags to the user, and (3) user selects from the pre-filtered options. This segmentation resolves the contradiction by breaking down the complex task of tag generation into manageable steps, where the system handles the complex analysis portion and the user handles only the simple selection portion
Data Source
AI summary
Embodiments of the present disclosure provide a solution for interaction. A method comprises: displaying a post interface of a media content; displaying a set of candidate tags in the post interface, the set of candidate tags being determined based on post information inputted through the post interface; and based on a selection of a target tag among the set of candidate tags, determining a target element corresponding to the selected target tag, the target element being configured to be displayed in a review page of the posted media content.