Cognitive Assistant for Predictive Content Discovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Enterprise File Sync and Share solutions burden users with the effort to find relevant content, as they require active participation and are not predictive of future interests, leading to duplication and inefficiency in content discovery.
Innovation Solution
An in-context cognitive information assistant that uses cognitive services to extract entities, topics, and concepts from user conversations and calendar activities, prioritizing relevant content for interactions and reducing the effort needed to find and share files.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If keyword search is used to find content, then the user can search for files by name or content characteristics, but the user must remember and enter specific search strings, increasing the effort and time required to find relevant content
Solution Approach 1:
The system performs preliminary indexing of file content, metadata, and user interactions before search is needed. Cognitive services pre-process and tag content with entities, topics, and concepts, so when a user searches, the system can quickly retrieve relevant results without requiring the user to formulate precise search strings or browse through results manually.
2Ease of operation
If folders and tagging are used to organize content, then files can be categorized and retrieved by category, but active manual involvement is required in advance to ensure files can be found later
Solution Approach 1:
The system enables self-service content organization by automatically analyzing file content, metadata, and user behavior patterns to generate tags, categories, and associations without requiring manual user input. Cognitive services automatically process uploaded content and assign appropriate metadata, allowing the system to organize itself based on actual usage patterns and content characteristics.
3Adaptability or versatility
If faceted browse is used to navigate content, then users can filter and browse content by multiple criteria, but active user participation is required and the process becomes laborious
Solution Approach 1:
The system incorporates feedback loops that monitor user interactions, search patterns, and content access behavior. This feedback is used to automatically adjust and refine content recommendations, tags, and associations. The system learns from user behavior and continuously improves its organization and retrieval mechanisms, reducing the need for manual filtering and browsing while increasing adaptability to user needs.
4Productivity
If recency search is used to find content, then files are retrieved based on historical user actions, but the system is not predictive of future interests and requires active user effort
Solution Approach 1:
The system performs preliminary analysis of user behavior patterns, content interactions, and contextual data to predict future information needs before users actively search. By pre-processing and pre-organizing content based on predicted requirements, the system can present relevant materials in advance, transforming reactive recency-based retrieval into proactive predictive delivery.
Solution Approach 2:
The system uses cognitive services to automatically analyze and understand user needs, content characteristics, and contextual relationships without requiring active user participation. This enables the system to self-adjust its content delivery based on predicted future interests, moving from passive response to active prediction and preparation of relevant materials.
Data Source
AI summary
An in-context cognitive information assistant is provided by: obtaining a context for a user, wherein the context comprises a calendar activity with one or more other users; supplementing the context by obtaining one or more conversations with the user related to the context; extracting cognitive data for the context and the conversations; and finding relevant materials in a corpus using the cognitive data. The relevant materials are used to prepare the user for interactions with other users.