Abstract:Monte Carlo tree search (MCTS) is one of the most capable online search algorithms for sequential planning tasks, with significant applications in areas such as resource allocation and transit planning. Despite its strong performance in real-world deployment, the inherent complexity of MCTS makes it challenging to understand for users without technical background. This paper considers the use of MCTS in transportation routing services, where the algorithm is integrated to develop optimized route plans. These plans are required to meet a range of constraints and requirements simultaneously, further complicating the task of explaining the algorithm's operation in real-world contexts. To address this critical research gap, we introduce a novel computation tree logic-based explainer for MCTS. Our framework begins by taking user-defined requirements and translating them into rigorous logic specifications through the use of language templates. Then, our explainer incorporates a logic verification and quantitative evaluation module that validates the states and actions traversed by the MCTS algorithm. The outcomes of this analysis are then rendered into human-readable descriptive text using a second set of language templates. The user satisfaction of our approach was assessed through a survey with 82 participants. The results indicated that our explanatory approach significantly outperforms other baselines in user preference.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the interpretability issue of the Monte Carlo Tree Search (MCTS) algorithm in sequential planning tasks, especially in traffic route services. Although MCTS performs well in fields such as resource allocation and traffic planning, its inherent complexity makes it difficult for users without a technical background to understand the operation process of the algorithm. Especially when multiple constraints and requirements need to be met, it becomes more difficult to explain the algorithm's choices to users.
To this end, the author proposes an interpreter based on Computation Tree Logic (CTL) to improve the transparency and interpretability of the MCTS algorithm. Specifically, this framework is implemented through the following steps:
1. **User - defined Requirement Transformation**: Use language templates to transform the user's query into strict logical specifications.
2. **Logical Verification and Quantitative Evaluation**: Check the states and actions traversed by the MCTS algorithm through the logical verification module to ensure that these states and actions conform to the user - defined logical specifications.
3. **Natural Language Generation**: Convert the results of verification and evaluation into human - readable descriptive text to help users understand the decision - making process of the algorithm.
### Main Contributions
- **Dynamically Associate User Queries**: Dynamically classify the queries of non - technical users and transform them into CTL formulas for analysis.
- **Utilize CTL Semantics**: Determine the satisfiability of logical queries and translate the results into natural language to ensure that the explanation of the algorithm is easy to understand and accessible.
- **User Satisfaction Evaluation**: Through a survey completed by 82 participants, the significant advantages of the CTL interpreter in terms of understandability, satisfaction, completeness, and reliability have been verified.
### Technical Details
- **Sequential Planning Model**: Use Markov Decision Process (MDP) to model public transport planning tasks and define the search space of the MCTS algorithm.
- **State and Action Definitions**: Define in detail the states and actions in the task, including vehicle location, request status, capacity limits, etc.
- **Reward Function**: Define a reward function that comprehensively considers factors such as the number of in - transit requests, the number of completed requests, and time efficiency.
### Application Scenarios
- **Dynamic Vehicle Scheduling**: In dynamic vehicle scheduling, the algorithm needs to process passenger requests in real - time, generate the optimal route plan, and simultaneously meet various constraints.
- **User Query Processing**: Users can pose various queries about the algorithm's decisions, such as "Why was this vehicle chosen instead of another?" or "Is this alternative feasible?"
Through these methods, the paper has successfully improved the interpretability and user acceptance of the MCTS algorithm in practical applications, especially in complex real - world scenarios such as public transportation.