18186120. MACHINE LEARNING SUMMARIZATION ON NON-STRUCTURED DIGITAL CONTENT simplified abstract (Microsoft Technology Licensing, LLC)
Contents
MACHINE LEARNING SUMMARIZATION ON NON-STRUCTURED DIGITAL CONTENT
Organization Name
Microsoft Technology Licensing, LLC
Inventor(s)
MACHINE LEARNING SUMMARIZATION ON NON-STRUCTURED DIGITAL CONTENT - A simplified explanation of the abstract
This abstract first appeared for US patent application 18186120 titled 'MACHINE LEARNING SUMMARIZATION ON NON-STRUCTURED DIGITAL CONTENT
The abstract describes a computerized method for summarizing digital content based on a user query. The method involves inferring the query to identify a website with non-structured content, extracting relevant media from the website, generating semantic summaries from the extracted content, and presenting an aggregation of the summaries to the user.
- Inference of user query to identify relevant website with non-structured content
- Extraction of most relevant media from the website based on the inference
- Generation of semantic summaries from the extracted content
- Aggregation of semantic summaries for presentation to the user
Potential Applications: - Content summarization for search engines - Automated content curation for websites - Personalized content recommendations for users
Problems Solved: - Efficient summarization of digital content - Improved user experience in accessing relevant information - Automation of content aggregation and presentation
Benefits: - Time-saving for users in finding relevant information - Enhanced user engagement with summarized content - Increased efficiency in content consumption
Commercial Applications: Title: Automated Content Summarization Technology for Enhanced User Experience This technology can be utilized by search engines, content platforms, and online media outlets to improve content discovery and user engagement. By automating the process of summarizing digital content, businesses can enhance the user experience and drive higher traffic to their platforms.
Questions about the technology: 1. How does the method infer user queries to identify relevant websites? 2. What are the potential challenges in generating accurate semantic summaries from extracted content?
Original Abstract Submitted
A computerized method for summarizing digital content based on a query from a user is described. An inference of the query is used to identify a website that includes non-structured content. The most relevant media within the website is identified based on the inference and content from the most relevant media is extracted. Using the inference, semantic summaries are generated from the extracted content, and an aggregation of the semantic summaries are presented to the user.
