Extracting Incomplete Planning Action Models from Unstructured Social Media Data to Support Decision Making

Abstract

Despite increasing interest in leveraging the wealth of online social media data to support data-based decision making, much work in this direction has focused on tasks with straightforward “labeling” decisions. A much richer class of tasks can benefit from the power of sequential decision making. However, supporting such tasks requires learning some form of action or decision models from unstructured data – a problem that had not received much attention. This paper leverages and extends machine learning techniques to learn decision models (incomplete action models) for planning from unstructured social media data. We provide evaluations showing the potential of unstructured data to build incomplete planning action models, which can further be extended to build PDDL-style action models for many real-world domains. Our models can be used to support novel quantitative analysis of online behaviors that can indirectly explain the offline behaviors of social media users