Decision Support for Knowledge Intensive Processes Using RL Based Recommendations

Supporting knowledge workers involved in the execution of unstructured Knowledge-Intensive Processes by providing context-specific recommendations remains an interesting challenge. Case data that represents expert decisions recorded in the past can be exploited for building a decision support tool for knowledge workers that can recommend which tasks to execute next. Reinforcement learning (RL) provides a framework for learning from interaction with the environment in order to achieve a certain process goal. RL has widely been used to model sequential decision problems and has shown great promise in solving large scale complex problems with long time horizons, partial observability, and high dimensionality of observation and action spaces [5]. In this paper, we propose a novel framework based on RL aimed at supporting knowledge workers by recommending the optimal course of action to the knowledge worker.