Semantic Text Mining of Organizational Processes - Discovering Whole-of-Government Collaboration Opportunities
Governments around the world pursue improvements in delivering public services, developing public policies, responding to crisis situations, and optimizing the use of public resources, among others. Achieving such goals must be based upon whole-of-government collaboration across different levels and functions of government, and across public and private sectors. In turn, establishing such collaboration requires information on the shared and individual goals, prospective participants, the roles to be filled and the resources to be contributed. Such information is rarely available as structured data (e.g. in databases) but instead scattered over portals, publications and other textual sources.
This research aims at investigating the use of semantic text mining for extracting collaboration-related information from unstructured data sources. Semantic text mining is a data mining and information extraction approach specialized in mining unstructured data sources using natural language processing and semantic technologies, such as ontologies and the semantic web. Specifically, we aim to exploit the online presence of prospective collaborating organizations, develop an ontology describing possible collaboration processes between them, and carry out a procedure analogous to process mining to extract shared, integrated or perhaps hybrid processes. These classes enable different collaboration scenarios: executing a single instance of a shared process on behalf of all organizations, executing integrated processes within such organizations as sub-processes of a larger multi-organizational process, or a combination of both approaches.
Expected contributions of this research are as follow:
- foundation - developing a novel approach to process-mining from unstructured data using semantic text mining,
- engineering - developing a tool, implementing the foundational approach, to aid the discovery of collaboration opportunities from websites and other forms of organizational online presence, and
- application - applying the tool for discovering collaboration opportunities in the whole-of-government context using real-life data.