Trends in BPM Research (Sommersemester 2020)

Dozent: Prof. Dr. Mathias Weske (Business Process Technology)

Allgemeine Information

• Semesterwochenstunden: 2
• ECTS: 3
• Benotet: Ja
• Einschreibefrist: 06.04.2020 - 22.04.2020
• Lehrform: Vorlesung
• Belegungsart: Wahlpflichtmodul
• Lehrsprache: Englisch

Studiengänge, Modulgruppen & Module

Beschreibung

Update (08.06.2020)

The sections on the exam and schedule were updated.

---

 

---

Business Process Management (BPM) concerns concepts and technologies to discover, model, execute, monitor, analyze, and improve processes within or between organizations. The Business Process Technology (BPT) group is engaged in versatile sub-topics of BPM, such as efficient use of resources in business processes, or the relationship between processes and smart contracts. Related fields like process mining or complex event processing are explored to exchange mutually beneficial concepts. Further, the challenges of inter-organizational interactions are addressed through process choreographies and blockchains.

The objective of the course is to present current research trends in BPM in general and at our chair in particular. We give insights into our daily work, during which you will also get to know techniques for approaching research life. This might support you in defining your own research topics, e.g., for your Master's thesis. Thus, this lecture is particularly interesting if you want to get to know or deepen your understanding of our chair and our projects.

Course Topics

Among others, the following topics will be discussed in this course.

Resource Management

What is a resource? In BPM, resources are an essential component and refer to mostly human resources (employees, roles) but also non-human resources (hard- and software components, time). Without resources, businesses could not execute their processes at all and an efficient use is directly correlated to the profit of the business. In BPMN, there are multiple concepts that help in allocating the right resource to the corresponding business step (activity): For human resources, there is the concept of lanes which represent the role, or the use of data objects and their attributes to model non-human resources. In Operation Research, there are resource allocation patterns that define different steps to successfully find a feasible resource-to-business-step allocation depending on the business case. However, these concepts only support a so-called “greedy” allocation logic that is easy to implement and understand, but severely limited in its efficiency. This part of the lecture focuses on the more advanced and complex allocation strategies which are not yet supported by existing business process modeling.

Process Mining

Data is everywhere nowadays, and organizations are striving for better techniques and methods to gain value from data. But how can data be used to provide insight on day to day operational processes? Process Mining is the discipline bringing data analysis to BPM. Process mining provides techniques to discover and analyze processes based on the real behavior, recorded in information systems. However, in order to exploit process mining capabilities, many steps have to be taken, from extracting and preparing data to performing various analysis techniques, to deploying results for process improvement and facilitating decision making. Each of these involved phases require efficient concepts as well as complementary technology. Process mining research concerns development and improvement of techniques to provide the right methods and tools for each step ensuring the realization of valuable result. We will also introduce our cooperation with Mount Sinai in New York, where we apply process mining techniques to real healthcare data.

Natural Language Processing

Automated techniques for the analysis of business processes provide a wide range of valuable opportunities for organizations. Among others, they allow to check for business process compliance, to identify redundant activities within an organization, and to identify operational overlap between two business processes. Traditionally, the input to such techniques are process models. That is, these techniques build on the formally specified relationships between the activities of process models to perform their analyses. This, however, means that these analysis techniques cannot be applied to less structured forms of process documentation such as textual process descriptions. Realizing that a considerable share of process documentation is based on natural language text in many organizations, this part of the lecture will focus on how Natural Language Processing techniques can be leveraged and applied in a process analysis context.

Blockchain and BPM

In this part of the lecture, we (i) learn about BPMN choreography models – a notation to describe the interactions in inter-organizational processes. Choreographies focus on the message exchange between collaborating parties and are executed without a central coordinator. Blockchain technology (especially in conjunction with smart contracts) can be used to control and coordinate a collaboration without centralizing it. Therefore we (ii) introduce blockchain technology. Using its distributed ledger, a blockchain is able to replicate a state in a peer-to-peer network. Smart contracts are programs that run transparently and tamper-proof on the blockchain. However, Blockchains bear some disadvantages. As a teaser for the next lecture we will have a brief discussion on blockchain-based choreographies and the technologies’ drawbacks.

We then introduce the notion of blockchain-based choreographies. This extension to BPMN choreography diagrams significantly enhances their expressiveness with a specific focus on blockchain capabilities such as shared data and logic. We show how these diagrams can be translated into smart contract code that enforces the choreography, including several pitfalls and limitations encountered. The lecture will close with a brief excursion into the field of legal informatics where we will discuss how smart contracts relate to actual legal contracts used in law today, and whether choreography diagrams might be a suitable abstraction to model them.

Voraussetzungen

There are no hard requirements for participating in this lecture, though some knowledge about modeling (e.g., POIS lecture) is certainly helpful.

Necessary background knowledge will be conveyed during the semester as far as possible.

Literatur

Mathias Weske: Business Process Management: Concepts, Languages, Architectures. Springer-Verlag Berlin Heidelberg 2019, 2012, 2007.

More literature will be published during the lecture by the respective lecturers.

Lern- und Lehrformen

The lecture will be held by the members of our chair. Each person will cover one or more lecture sessions.

Leistungserfassung

There will be oral exams on 14.07, 15.07, and 16.07.

The exact format is not yet fixed due to COVID-19, but the following is likely:

• The exams will be held in groups of two students, and will last around 30 minutes.
• Time slots between 9am and 11am on each day will be distributed according to preferences.

Termine

Week	Date	Topic	Presenter
18	27.04	Introduction & Course Structure	Jan Ladleif (Zoom)
19	04.05	Business Process Management	Mathias Weske (OpenHPI)
20	11.05	Decision Mangement	Mathias Weske (OpenHPI)
21	18.05	Resource Management	Sven Ihde
22	25.05	Case Management & Cross-Case Dependencies	Stephan Haarmann
23	01.06	(Pfingstmontag, Pentecost)	---
24	08.06	Natural Language Processing in BPM	Henrik Leopold
25	15.06	Process Mining & Data Model Discovery	Dorina Bano
26	22.06	Event Log Generation	Simon Remy
26	25.06 (Thursday)	Mining Resource Behavior	Maximilian Völker
27	29.06	Blockchain and BPM	Jan Ladleif
28	06.07	Blockchain Architectures	Ingo Weber, TU Berlin
28	09.07 (Thursday)	Course Summary	Jan Ladleif

Zurück