Scene and Chunk Estimation for Targeted Work Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing systems require large-scale reconstruction of documents to provide relevant information, leading to unnecessary information overload and inefficient response times.
Innovation Solution
An information processing device that estimates scenes and chunks using trained models to selectively output work-related information based on acquired images, location, and time, without reconstructing large amounts of data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If documents are reconstructed into structured information to enable detailed searching, then information retrieval precision is improved, but device complexity and processing cost increase significantly
Solution Approach 1:
The patent segments the document processing task by extracting only relevant features (scene ID, chunk ID, chunk-meta ID) rather than reconstructing entire documents. This segmentation allows precise information retrieval through feature matching while avoiding the complexity of full document structuring.
Solution Approach 2:
The patent extracts essential identifying features (scene ID, chunk ID, chunk-meta ID) from documents without performing complete document reconstruction. This extraction approach enables targeted information retrieval while significantly reducing processing complexity compared to full document structuring.
2Quantity of substance
If all documents are provided to the user to ensure complete information availability, then information completeness is improved, but information overload increases and response time decreases
Solution Approach 1:
The patent applies local quality by providing different information quantities to different users based on their specific needs and contexts. Instead of uniform information delivery, the system tailors information presentation to individual requirements, ensuring completeness where needed while avoiding overload elsewhere.
Solution Approach 2:
The patent implements partial action by delivering only the necessary portion of information required for each specific task or query. Rather than providing all documents universally, the system provides partial information sets tailored to specific needs, reducing overload while maintaining necessary completeness.
3Quantity of substance
If document information is presented in document units, then information completeness is maintained, but information relevance decreases due to inclusion of unnecessary information
Solution Approach 1:
The patent extracts and presents only the relevant portions of information (specific chunks identified by chunk ID and chunk-meta ID) rather than presenting entire documents. This extraction maintains necessary information completeness while eliminating unnecessary content, thereby improving relevance.
Solution Approach 2:
The patent segments information at the chunk level rather than document level, allowing selective presentation of relevant segments. This segmentation enables the system to maintain information completeness for required content while excluding unrelated portions, improving overall relevance.
Data Source
AI summary
An information processing device to output work information, which is information related to work performed by a user, has a scene estimation unit to acquire a first image, location information, and time information, all relating to the work, and to estimate a scene by using a first trained model, a chunk estimation unit to acquire a second image, of equipment related to the work, and estimate a chunk by using one of a plurality of second trained models, in which associations among the second image, a chunk ID that uniquely identifies the chunk, which is information that divides or suggests the work information, and one or a plurality of chunk-meta IDs that are associated on a one-to-one basis, are stored, and an output unit to output the chunk.