Image Management Server with Dual Models for Tacit Knowledge Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to effectively convey tacit knowledge across different tasks and between users with varying levels of expertise, particularly in industries like civil engineering and construction, due to the qualitative nature of tacit knowledge and challenges in ensuring user trust and relevance.
Innovation Solution
An information processing system comprising an image management server and terminal device, which utilizes first and second tacit knowledge models trained on correspondences between image information and input information, generating text information based on selected image information and user input, and transmitting this information to the terminal device for display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single tacit knowledge model is used for all tasks and users, then device complexity is reduced, but the relevance and trust of generated text information deteriorates for users with varying expertise levels
Solution Approach 1:
The patent divides the single tacit knowledge model into multiple specialized models (first tacit knowledge model and second tacit knowledge model) with different knowledge granularities. The first model contains detailed operational expertise while the second model provides general-purpose knowledge. This segmentation allows the system to match appropriate models to user needs, improving trust in generated information without requiring a completely separate system for each user type.
Solution Approach 2:
The patent implements dynamic model selection where the system automatically chooses between the first and second tacit knowledge models based on the user's expertise level and task requirements. This dynamic adaptation allows a single interface to serve multiple user types effectively, maintaining simplicity while improving reliability through contextual model matching.
2Measurement precision
If detailed operational expertise is included in all text information, then the quality of information for expert users improves, but the accessibility and relevance for novice users deteriorates
Solution Approach 1:
The patent applies local quality by creating different levels of knowledge detail within the model system. The first tacit knowledge model provides highly detailed operational expertise for specific tasks, while the second model provides simplified general knowledge. This allows the system to deliver appropriately detailed information to each user type without overwhelming novices with expert-level complexity or under-servicing experts with insufficient detail.
3Reliability
If multiple specialized models are maintained for different user types, then the relevance of text information improves, but device complexity and model management burden increases
Solution Approach 1:
The patent implements self-service through automatic model selection based on user expertise level detection. The system autonomously determines which model to use (first or second tacit knowledge model) without requiring manual user input or complex configuration. This automation handles the model management complexity internally while maintaining high relevance for each user type.
Solution Approach 2:
The patent creates a universal interface that handles multiple user types through a single unified system architecture. Both expert and novice users interact with the same interface, and the system automatically adapts the underlying model selection. This multi-functionality approach allows the system to serve diverse needs without requiring separate specialized interfaces for each user type.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing system (100) includes an image management server (40) and a terminal device (10). The image management server (40) manages image information of an item. The terminal device (10) communicates with the image management server (40). The image management server (40) includes a memory (4000) and a text information generation unit (45). The memory (4000) stores a first model (4004A) and a second model (4004B). The first model (4004A) is trained on a correspondence between the image information of the item and information on the item, or a correspondence among the image information of the item, the information on the item, and input information input into the terminal device (10). The second model (4004B) is trained on a correspondence between the image information of the item and the input information input into the terminal device (10). The text information generation unit (45) generates text information on the item based on the image information of the item selected via the terminal device (10), the information on the item, and the first model (4004A), or on the image information of the item selected via the terminal device (10) and the second model (4004B). The terminal device (10) includes a display control unit (13) that causes a display screen including the text information to be displayed.