Self-learning Predictor Aggregation for the Evolution of People-Driven Ad-Hoc Processes

Abstract

Abstract. Contemporary organisational processes evolve with people's skills and changing business environments. For instance, process documents vary with respect to their structure and occurrence in the process. Supporting users in such settings requires sophisticated learning mechanisms using a range of inputs overlooked by current dynamic process systems. We argue that analysing a document's semantics is of uttermost importance to identify the most appropriate activity which should be carried out next. For a system to reliably recommend the next steps suitable for its user, it should consider both the process structure and the involved documents' semantics. Here we propose a self-learning mechanism which dynamically aggregates a process-based document prediction with a semantic analysis of documents. We present a set of experiments testing the prediction accuracy of the approaches individually then compare them with the aggregated mechanism showing a better accuracy.

Enabling the Analysis of Cross-Cutting Aspects in Ad-Hoc Processes

“…Some theoretical works [6,5] explored declarative approaches to specifying the artifact lifecycles following an event oriented style. Another line of work in this category, focused on modeling and querying artifact-centric processes [20,30,11]. In [20,30], a document-driven framework, proposed to model business process management systems through monitoring the lifecycle of a document.…”

“…Dorn et.al. [11], presented a self-learning mechanism for determining document types in peopledriven ad-hoc processes through combining process information and document alignment. Unlike our approach, these approaches assumed a predefined document structure or they presume that the execution of the business processes is achieved through a BPM system (e.g., BPEL) or a workflow process.…”

Dynamic Collective Work [Guest editors' introduction]

“…Industries such as healthcare, insurance, and banking have witnessed explosive growth in such processes. [5][6][7][8][9][10][11][12][13] Dynamic and collective activities are characterized by their flexibility and data-driven nature. Automobile insurance claims processing, order processing of prescription drugs, patient case management in a hospital, and recovery and response assistance during natural disasters are a few examples.…”

“…T he articles in this issue make it clear that analyzing and using dynamic collective work activities is increasingly relevant today. Software professionals, founders and employees of start-up companies, researchers, and social scientists are actively participating in this space and exploring solutions to problems such as designing recommender systems, [6][7][8] predicting next steps for case workers, 9 analyzing and tracing healthcare processes, 10,11 adapting mined models to dynamic changes in a process, 15 and optimizing business processes in crowdsourcing platforms. 12,13 Analysis of dynamic collective work could provide this spectrum of users with insight for making better decisions and introduce process optimizations.…”