E-commerce Recommendation Words for Multimedia Search Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face challenges in finding desired information when browsing multimedia content, as they can only see limited information directly from the content, leading to a long and inefficient search process with a low target achievement rate.
Innovation Solution
An information recommendation method and apparatus that determines a product included in multimedia content displayed on a first page, generates e-commerce recommendation words based on the product and e-commerce attribute features, and displays these words on a second page, allowing users to initiate a targeted search.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users search based on limited directly observed information from multimedia content, then the search process becomes longer and less efficient, but users can only see limited information directly from the content
Solution Approach 1:
The system performs preliminary action by automatically extracting product information and generating e-commerce recommendation words before the user initiates a search. The recommendation words are pre-generated based on product attributes and displayed on the search page, so when the user triggers a search, the relevant search terms are already prepared and suggested, significantly reducing the search time and improving efficiency.
Solution Approach 2:
The patent introduces e-commerce recommendation words as an intermediary element between the multimedia content and the search query. These recommendation words serve as a bridge that translates the limited visible information from the content into comprehensive search terms, helping users expand their search scope without manually analyzing all available information.
2Reliability
If users try various possible search information to find desired content, then the target achievement rate improves, but the search process becomes longer
Solution Approach 1:
The system pre-generates multiple e-commerce recommendation words based on product attributes before the user needs to search. These recommendation words are displayed on the search page, allowing users to directly trigger searches with pre-prepared terms that are likely to achieve their targets, thereby reducing the time needed to formulate effective search queries.
Solution Approach 2:
The system provides self-service by automatically generating and presenting relevant search terms based on the multimedia content. Instead of requiring users to manually analyze and formulate search queries, the system serves itself by extracting product information and generating appropriate search terms, which users can then directly use to initiate searches with higher target achievement rates.
Data Source
AI summary
The present disclosure provides an information recommendation method and apparatus, a computer device, and a storage medium. The method comprises: determining, in response to a search instruction for a first page, a product included in a multimedia content displayed on the first page; displaying, on a second page, at least one e-commerce recommendation word, the e-commerce recommendation word being generated based on the product and a keyword corresponding to at least one e-commerce attribute feature; receiving a trigger operation for the e-commerce recommendation word, and initiating, based on the e-commerce recommendation word and a vertical domain of content corresponding to the e-commerce recommendation word, a search request.