Interactive Content Construction via DAG Planning Domain
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for constructing interactive contents, such as the branch and planning schemes, require significant time and effort to predict user interactions and intuitively construct procedures to achieve a goal state, as they do not allow for flexible and efficient reuse of interactive content data.
Innovation Solution
A data processing apparatus and method utilizing a directed acyclic graph (DAG) to receive, analyze, and generate training data, which is then used to construct a planning domain with characterized plan operators and pre- and post-conditions, enabling flexible construction and reuse of interactive content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the branch scheme is used to construct interactive contents, then all user interactions can be predicted and set in advance, but the device complexity and time required to set up all branches increases significantly
Solution Approach 1:
The system uses machine learning models that automatically learn user interaction patterns from training data, eliminating the need for manual prediction and setup of all branches. The model self-adjusts to predict user actions at each node based on learned patterns, reducing the complexity of pre-defining all possible branches while maintaining reliable interaction prediction.
2Productivity
If the planning scheme is used to construct interactive contents, then procedures can be automatically constructed, but the procedures are not intuitive and significant time and effort are required to achieve the goal state
Solution Approach 1:
The system performs preliminary training by learning from training data containing user interaction patterns before actual content construction. This pre-learning phase enables the model to quickly generate intuitive procedures during runtime, reducing the time required to achieve goal states while maintaining automatic construction capabilities.
Solution Approach 2:
The system changes the approach from rule-based procedure construction to probabilistic prediction based on learned patterns. By adjusting parameters such as prediction confidence thresholds and using learned interaction patterns instead of fixed rules, the system generates more intuitive procedures that require less time to execute.
3Ease of manufacture
If traditional methods are used for constructing interactive contents, then contents can be created, but flexible and efficient reuse of interactive content data is not allowed
Solution Approach 1:
The system uses a universal machine learning model that can be trained on diverse interactive content data and applied across different content types. The learned patterns and features are reusable across multiple projects and content constructions, enabling flexible and efficient reuse of interactive content data while maintaining ease of content creation.
Data Source
AI summary
Disclosed are an apparatus and method of constructing interactive contents. The data processing apparatus may transit a branching graph into a planning domain, wherein information on nodes and arcs of the branching graph may be classified to generate property and class information, and the classified properties may be used as pre-conditions of a plan operator on the planning domain.