Video Classification Using Humanistic Attributes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video search engines rely on external labels and fail to accurately reflect video content, leading to low matching results due to user-specific value judgments and background differences.
Innovation Solution
A method and apparatus for multi-dimensional video content identification to generate humanistic attribute information, enabling more accurate video classification and recommendation based on user values and usage scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video search engines use external labels (title, state photo) for classification, then the search process is simple and fast, but the matching accuracy with video content is low
Solution Approach 1:
The patent segments video classification into multiple dimensions: external label classification (title, state photo) and internal content classification (abstract information, humanistic attributes). This segmentation allows the system to maintain simple external search interfaces while implementing complex multi-dimensional content analysis internally, resolving the contradiction between simplicity and accuracy.
Solution Approach 2:
The patent adds new classification dimensions beyond traditional external labels by extracting abstract information and humanistic attributes from video content. This dimensional expansion transforms the classification system from two-dimensional (external labels only) to multi-dimensional (external labels + content abstract + humanistic attributes), significantly improving matching accuracy while managing complexity through structured processing.
2Loss of information
If video classification is based only on external labels, then the system is easy to implement, but it cannot reflect the true content information of videos
Solution Approach 1:
The patent performs preliminary extraction of abstract information and humanistic attributes from video content during the classification process. By pre-processing and structuring the content information extraction, the system reduces the complexity of real-time content analysis while ensuring comprehensive content information is captured and stored for accurate matching.
3Measurement precision
If the system considers user-specific value judgments and background differences, then personalized recommendation accuracy improves, but the complexity of user profiling and matching increases
Solution Approach 1:
The patent applies local quality by differentiating between universal video attributes (external labels, abstract information) and user-specific attributes (value judgments, background preferences). The classification system maintains standardized processing for video content while allowing personalized matching algorithms to operate on user-specific dimensions, reducing overall system complexity while enabling accurate personalization.
Data Source
AI summary
The object of the present disclosure is to provide a method, apparatus and selection engine for classification matching of videos. The method according to the present disclosure includes: performing multi-dimensional identification of the content of at least one video in order to determine the abstract information of the at least one video; generating the respective classification attribute information of the at least one video based on the respective abstract information of the at least one video, wherein the classification attribute information includes the humanistic attribute information of the video; wherein the humanistic attribute information is used to indicate the value judgement corresponding to the video. The present disclosure has the following advantages: fully excavate various classification attributes of video content through multi-dimensional identification of the video, and then obtains the humanistic attributes of the video. Therefore, it can be sifted based on the humanistic attributes of the video and be recommended to users that are more closely matched.