Intelligent Agents System for Intention Mining Using HMM-LSTM Model

Abstract

With the significant increase in information in the system log files, much research is based on the log files to improve the process. System logs contain a lot of information at different times about the behavior of the system and user. To enhance the process based on log files, the users’ behavior must be learned. With process discovery, an intentional process model and prediction strategies can be learned from log data. However, system logs do not fulfill the requirements that process discovery and prediction algorithms place on log files. To solve this problem, a new ensemble-based multi-intelligent agent system is introduced for discovering intentional process models and predicting users’ strategies in intention mining. This research proposal, therefore, proposes an architecture of four layers for intelligent agents to generate an intention mining process based on the communication coordination of intelligent agents, and we propose an HMM-LSTM-based hybrid solution to model and predict strategies of students using the case study of Educational Process Mining (EPM): A Learning Analytics Data Set.