Scholarlink

Constraint Programming (CP) serves as a widely used platform in both industry and research, simplifying the modeling and resolution of complex constraint satisfaction and optimization problems. However, the automated model reformulation platforms that aims to assist the users in modeling and solving constraint problems remains restricted. This level of restriction is primarily caused by the mathematical modeling and programming skills required to handle various types and forms of Nondeterministic Polynomial Complete (NP-complete) decision problems, including challenges related to planning, scheduling, and resource optimization. In this context, we present our research, which aims to integrate Natural Language Processing (NLP) with constraint acquisition via a conversational agent. The objective is to enhance the intuitiveness of the constraint programming paradigm, enabling a more diverse range of users to interact with it effectively. The study focuses on the development of an intelligent constraint acquisition system that allows novice users to communicate their problem in natural language. The system interprets the provided information, formulates the given problem as a constraint network, solves it, and presents alternate solutions to the user when available.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning and NLP-Based Approach for Constraint Acquisition Problems

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Abdessalam Bahafid,
  • Zakarya Erraji,
  • Naira Abdou Mohamed,
  • Anass Allak,
  • Kamel Gaanoun,
  • Imade Benelallam

摘要

Constraint Programming (CP) serves as a widely used platform in both industry and research, simplifying the modeling and resolution of complex constraint satisfaction and optimization problems. However, the automated model reformulation platforms that aims to assist the users in modeling and solving constraint problems remains restricted. This level of restriction is primarily caused by the mathematical modeling and programming skills required to handle various types and forms of Nondeterministic Polynomial Complete (NP-complete) decision problems, including challenges related to planning, scheduling, and resource optimization. In this context, we present our research, which aims to integrate Natural Language Processing (NLP) with constraint acquisition via a conversational agent. The objective is to enhance the intuitiveness of the constraint programming paradigm, enabling a more diverse range of users to interact with it effectively. The study focuses on the development of an intelligent constraint acquisition system that allows novice users to communicate their problem in natural language. The system interprets the provided information, formulates the given problem as a constraint network, solves it, and presents alternate solutions to the user when available.