AUTOMATED KNOWLEDGE BASE CONSTRUCTION USING PROCESS LOGS

The concept of building and using the knowledge base in the process management system corresponds to the Enterprise 2.0 Management Paradigm. The key difference of this paradigm is the integration of the collective work of the knowledge workers into the existing business processes. Both formal and informal knowledge are used when integrating. This makes it worthwhile to combine the benefits of process and functional management approaches. The combination of these approaches leads to the creation of flexible knowledge-intensive business processes at the enterprise [1]. When executing these processes, knowledge workers can make decisions that change the workflow in accordance with the current tasks, taking into account the state of the subject area on the basis of formal and implicit personal knowledge. The implementation of this paradigm is based on the integration of process management and knowledge management. In general, when solving these problems, it is necessary to transform the personal tacit knowledge into the explicit form [2], to ensure the storage of this knowledge and their use by the workers [3].

Thus, the actual task is to develop an approach to the automated construction of the knowledge base of the process management system based on the analysis of the behavior of business processes.

Knowledge Base Construction (KBC) is a process of creating knowledge bases by identifying facts and relationships between them in the input [4].

An analysis of the approaches to automated KBC confirms that knowledge bases should have semantic and informational components [5, 6]. The first component provides knowledge representation, and the second is the connection with the input information, on the basis of which the presentation of knowledge is constructed. According to our approach, integration of these two aspects takes place on the basis of events, since the events are recorded in the log file and reflect the execution of the process in a concrete context.

An event in the process control system will be considered as a change in the state of the process, which is characterized by the following properties: by implemented action, or change in the state of action that is being performed; temporal aspect in the form of a timestamp, when the event occurred; a spatial aspect that reflects the state of the process's context when the event occurred.

In order to provide a complete representation of the knowledge of an event, we need to combine information from different tracks and possibly from several logs of similar processes. Such information may be controversial, depending on the sources of input data. However, the comparative analysis of the contradictions in the description of the events enables one of the research tasks to be solved - to identify and justify new sequences of actions performed on the basis of the personal experience of the performers. Then, new actions, rules and limitations on their use can added to the Knowledge Base and an improved process model.

Thus, the knowledge base of the process information system should contain the following elements that are represented on the semantic and informational levels: events, artifacts, relations. A number of classes of artifacts and events are proposed. Relation is discussed as control rules and restriction.

There is knowledge of the relationship between artifacts and business process events at the semantic level. The information level contains mentions to these events, links, and artifacts in the logs.

The construction of the knowledge base has the following sequence: the core is first formed using known explicit knowledge. Then, using log analysis, unique events and artifacts, and relationships between them are determined. These relationships specify the rules and limitations for implementing the business process in spatial and temporal aspects.

The spatial aspect is described as a context of the business process. The sequence of works implements the temporal aspect of the business process

References:

1. Gronau, N. (2012). Modeling and Analyzing knowledge intensive business processes with KMDL: Comprehensive insights into theory and practice (English). Gito, 522.

2. Polanyi, M. (1958) Personal Knowledge: Towards a Post-Critical Philosophy. University of Chicago Press, 493.

3. G. Alvaro, C. C´ordoba, V. Penela, M. Castagnone, F. Carbone, J.M. G´omez-P´erez, J. Contreras (2011) An intra-enterprise semantic microblogging tool as a micro-knowledge management solution. ESWC 2011: The Semanic Web: Research and Applications, 154-168.

4. Fabian M. Suchanek, James Fan, Raphael Hoffmann, Sebastian Riedel, Partha Talukdar (March, 2013) Advances in Automated Knowledge Base Construction. SIGMOD Records.

5. Jaeho Shin, Sen Wu, Feiran Wang, Christopher De Sa, Ce Zhang, Christopher Ré Kasneci G. (2015) Incremental Knowledge Base Construction Using DeepDive. VLDB Endowment. Vol.8(11), 1310 – 1321.