Abstract
The employment of batch processing in workflow is to model and enact the batch processing logic for multiple cases of a workflow in order to optimize business processes execution dynamically. Our previous work has preliminarily investigated the model and its implementation. However, it does not figure out precisely which activity and how a/multiple workflow activity(s) can gain execution efficiency from batch processing. Inspired by workflow mining and functional dependency inference, this chapter proposes a method for mining batch processing patterns in workflows from process dataflow logs. We first introduce a new concept, batch dependency, which is a specific type of functional dependency in database. The theoretical foundation of batch dependency as well as its mining algorithms is analyzed and investigated. Based on batch dependency and its discovery technique, the activities meriting batch processing and their batch processing features are identified. With the batch processing features discovered, the batch processing areas in workflow are recognized then. Finally, an experiment is demonstrated to show the effectiveness of our method.
Top2. Problem Definition
In this section we will give a brief introduction to what batch processing in workflow is, what data we will employ and how mining steps start and proceed.
Key Terms in this Chapter
Workflow mining: Workflow mining is a novel modelling method, which can automatically derive workflow model from workflow logs or from transaction logs of information system using transactional systems such as ERP, CRM.
Workflow: A workflow is a partial or total automation of a business process. More abstractly, a workflow is a pattern of activity enabled by a systematic organization of resources, defined roles and mass, energy and information flows, into a work process that can be documented and learned.
Batch processing area: A batch processing area is a set of sequence ordering activities in workflow, in which activities have common batch processing feature and sequence ordering relation holds between them.
Batch dependency: A batch dependency is a constraint of a relation, which holds if there is a pair of attribute or attribute set of the relation satisfies some conditions.
Batch processing feature: A batch processing feature is one or a set of input parameters of an activity in workflow, on which activities can be batch-processed.
Data Mining: Data mining is the nontrivial extraction of implicit, previously unknown, and potentially useful information from data. It is a science of extracting useful information from large data sets or databases.
Functional dependency: Functional dependencies are relationship between attributes of a database relation: a functional dependency states that the value of an attribute is uniquely determined by the values of some other attributes.