English teaching data often contains a amount of textual information, which poses certain difficulties for teaching due to its large volume, complex types, and difficulty in efficient localization. Therefore, a method of English teaching data location based on multi-agent hierarchical reinforcement learning is designed. Extract the challenge response delay monomer feature of English teaching data, use machine learning and data analysis to process the key information of English teaching data, and feedback the data type under the hierarchical reinforcement learning framework. A positioning model for English teaching data is formulated utilizing a multi-agent approach within a hierarchical reinforcement learning framework. Based on the characteristics and requirements of English teaching data, an appropriate multi-agent system architecture is designed, and the number, type and function of agents are determined, so as to allocate multiple subtasks to automatically locate teaching data. Generate topological adaptive levels of English teaching data positioning, dynamically adjust the hierarchical relationship, interaction mode and communication protocol of agents according to students’ learning progress and feedback, so as to meet the data positioning requirements of different teaching environments. The research results demonstrate that the proposed method exhibits greater efficiency.