sebis research profile

Software Engineering für betriebliche Informationssysteme (sebis) Fakultät für Informatik Technische Universität München wwwmatthes.in.tum.de

sebis Research Profile 20.7.2014, Prof. Dr. Florian Matthes

Research background

© sebis Sebis Research Profile 2

Enterprise Architecture Management

Social Software Engineering

§  System cartography §  EAM tool surveys §  EAM pattern catalog §  Capability models in

mergers & acquisitions §  Building blocks for EAM §  Wiki4EAM §  Agile EAM

§  User-centered social software

§  Authorization models in social software

§  Introspective model-driven development

§  Enterprise 2.0 tool surveys §  Hybrid Wikis §  Tag-based knowledge

organization

Communities

Collaborative Work

Digital Content

§  CoreMedia AG (Spinoff) §  infoAsset AG (Spinoff) §  Business & IT

transformation @ VW §  EAM 2.0 @ HUK Coburg §  KPI systems @ SFS §  Cloud security @ Siemens §  Strategy assessment @ FI §  D-MOVE

Technology Transfer Projects

more >

Team


more >

Alexander Schneider

Matheus Hauder

Klym Shumaiev

Thomas Reschenhofer

Marin Zec

Florian Matthes

Aline Schmidt

Jian Kong

Enterprise Architecture Management

Social Software Engineering

Bernhard Waltl

Alexander Waldmann

Project partners since 2002


Enterprises and public administrations

Deutsche Börse Systems

Project partners since 2002


Consultants and software vendors

Academic education


Bachelor Informatics

§  Introduction to Software Engineering

§  Software Engineering for Business Applications

§  Software Engineering in Industry and Practice

Master Informatics

§  Strategic IT Management and EAM

§  Web Application Engineering

§  Software Architectures §  Global Software

Engineering §  GFSU (Startups,

Entrepreneurship)

Life-Long Learning

§  Euro CIO Professional Programme in Business and Enterprise Architecture

§  EAMKON Conference Series

§  Softwareforen Leipzig Working Group EAM

more >

Prototypical Solutions

Practical Experience

Informatics Models

Information & Communication

Technology

Research approach

© sebis Sebis Research Profile

Abstraction

Application

Evaluation Engineering

Spin-Off

7

Informatics Application Domain

Research projects and results

1.  Enterprise Architecture Management §  IT Architecture in Turbulent Times §  Agile Enterprise Architecture Management §  Quantitative Models in Enterprise Architecture Management §  Federated Enterprise Architecture Model Management §  CALM3: Complexity of Application Landscapes §  Semantic Processing of Legal Texts for IT Compliance

2.  Social Software Engineering §  Darwin: Process Support for Collaborative Knowledge Work §  Spreadsheets 2.0: Analysis of Complex Linked Data §  Social Software for Complex Problem Solving §  COLVA: Collaborative Learning Video Annotations


The adoption rate for new technologies keeps accelerating.


Forbes Magazine July 7th 1997

Exponential growth starts inconspicuously, and humans are not used to reasoning about non-linear processes.

Sebis Research Profile 10 © sebis

Google Trends December 2013

Humans: Employees, Customers, Suppliers, Partners, Markets, Communities, … Laws & Regulations

Resources: Energy, Matter, Information, Technology…

Enterprise

An enterprises understood as an adaptive system of systems


Business Capabilities

Information Management

OPTIMIZE TRANSFORM

IM Capabilities

OPTIMIZE TRANSFORM

Goals, Strategy

Vision, Goals, Strategy

Motivation – Most frequent EA challenges

0,00%

10,00%

20,00%

30,00%

40,00%

50,00%

60,00%

70,00%

80,00%

90,00%

100,00%

1. Ad hoc EAM demands

2. Unclear business goals

3. Hard to find experienced

enterprise architects

4. EA demands unclear for EAM

team

5. Enterprise environment

changes too quickly

Agree (%)

Neither (%)

Disagree (%)

n=102

13

Hauder, M., Roth, S., Schulz, C., Matthes, F.: Organizational Factors Influencing Enterprise Architecture Management Challenges, 21st European Conference on Information Systems (ECIS 2013), Utrecht, Netherland, 2013.

© sebis Sebis Research Profile

Agile EA management principles

Sebis Research Profile 14

Individuals and interactions over formal processes and tools

IT Project 3 IT Project 2 IT Project 1

Top management

Business stakeholders

Software development

IT operations

Project managers

Software architects

Software developers

Top management

Strategy office

Business owners

Application owners

IT operations

Purchasing

EA Team

•  Ensure top management support

•  Maintain a good relationship to people form other management areas

© sebis



Focus on demands of top stakeholders and speak their languages


Architecture blueprints

Top management



IT operations

Project managers

Software architects

Software developers

communicate

explain

involve

support

get feedback

� �

Top management

Strategy office

Visualizations Business owners

Application owners

IT operations

Purchasing

EA Team

Stakeholder-specific architecture views

Metrics

Reports

Architecture- approval and requirements

Architecture changes

model

collect

motivate

Business and IT strategy

Individual architecture aspects

Business and org. constraints

•  A single number or picture is more helpful than 1000 reports

•  Communicate, communicate, communicate

•  Avoid waste •  Benefit form existing model

management processes

© sebis



Reflect behavior and adapt to changes


Architecture blueprints

Top management



IT operations

Project managers

Software architects

Software developers

�

communicate

explain

involve

support

get feedback reflect

adapt

� �

Top management

Strategy office

Visualizations Business owners

Application owners

IT operations

Purchasing

EA Team

Stakeholder-specific architecture views

Metrics

Reports

Architecture- approval and requirements

Architecture changes

model

collect

motivate

Business and IT strategy

Individual architecture aspects

Business and org. constraints

© sebis

•  Iterative and Incremental (one cycle ~12 months)

•  Use building blocks and patterns

•  Request 360° feedback •  Adapt models and processes •  Continuous collaboration

Using quantitative models in the context of EAM


System structure (EA, static)

change change

con

stra

ins

𝒕−𝟏 t = 𝑵𝑶𝑾

𝒕+𝟏

1 3

System behavior (dynamic)

4

1.  Assess the

architecture with metrics

2.  Measure architecture changes

3.  Plan architecture changes

4.  Monitor system performance with KPIs (Business & IT)

2

con

stra

ins

con

stra

ins

Metric Management Method (MMM) as Extension of the BEAMS Conceptual Framework


Stakeholders

Goals + Concerns

Organizational context

Organizational context

Organizational Context

Actors Enterprise Architects Enterprise Architects

Development method

Characterize situation Configure EAM function Analyze EAM function

Adapt and evolve EAM function

Execute EAM

function

Implementation Guide (Patterns & Building Blocks)

BEAMS , EAM Pattern Catalog and EAM KPI Catalog

EA Metric

VBB

Performance Indicator

VBB VBB

IBB

EA Metric

IBB IBB

+ EAM Metric Catalog

Integrated software support for quantitative models in the domain of EAM

Best practices for EAM metrics & performance measurement §  KPI template §  KPI catalog §  Method for designing a KPI system

Integrated Software Support

§  Query language for KPI definition over complex information models §  KPI visualization (in progress)

Evaluation

§  Siemens Financial Services §  Credit Suisse, Bayern LB, Commerzbank, CALM3


Type of collection n % of all Manually from applications/databases 95 76.00%

Manually via interviews 85 68.00%

Manually modeled in workshops 66 52.80%

Manually via questionnaires 46 36.80%

Partially collected automatically 44 35.20%

What are current problems in EA model maintenance?


N=125, 2013

Challenge n % of all Huge data collection effort

77 55.00%

Low EA model data quality 77 55.00%

Insufficient tool support 48 34.29%

No management support 44 33.43%

Low return on investment 36 25.71%

Other 32 22.86%

No specific challenge 10 7.14%

More >

Federated enterprise architecture model management


Modeling communities, artifacts, processes and their interactions

E

Metamodel and Model

D EAM

Task

Technology

fit

Metamodel and Model

A PPM

Task

Technology

fit

Team

publish model changes

model and meta-model changes to be

integrated

Metamodel and Model

B BPM

Task

Technology

fit

Team


Metamodel and Model

C ITSM

Task

Technology

fit

Team



Federated EA Model Management

Modeling Experts Modeling Community Metamodel Mappings Instance Mappings

Team

Enterprise

•  Importing •  Differencing •  Conflict detection •  Conflict resolution

•  Collaboration •  Negotiation

Federated enterprise architecture model management

1.  Import of different models in a metamodel-based EA tool

2.  Synchronization via model merging Provide means to identify model elements within the originating information source

3.  Conflict detection during merge operation §  Instance conflicts §  Schema conflicts §  Schema/instance conflicts

4.  Collaborative conflict resolution Fine-grained access control is employed to find the organizational role in a chain of responsibility

5.  Customizable conflict resolution strategy


Tool support - ModelGlue

For further information see https://wwwmatthes.in.tum.de/pages/kkdtsjtjkc2g

CALM3: Complexity of application landscapes

Research questions §  What does "IT-complexity“ mean? §  How can complexity be described? §  Which factors drive application landscape complexity? §  How can complexity be quantified? §  How can complexity models contribute to landscape

planning?


Models, metrics and methods

Project Partners

CALM3 Workshop

Series

10 Industry experts

Quarterly meetings

Extensive EA data

Concrete metrics

Tool development

Visionary discussions

The complexity cube


The complexity cube


Classifying EA literature

EA Complexity Publications

ACN D1 ACN D2 ACN D3 ACN D4

Janssen et al. (2006) qualitative structural, dynamic objective ordered

Buckl et al. (2009) qualitative structural objective ordered

Saat et al. (2009) qualitative structural, dynamic objective ordered

Dern et al. (2009) quantitative structural objective disordered

Mocker (2009) quantitative structural objective disordered

Zadeh et al. (2012) qualitative, quantitative structural objective ordered

Kandjani et al. (2012) quantitative structural objective ordered

Kandjani et al. (2013) qualitative, quantitative dynamic objective ordered

Schütz et al. (2013) quantitative structural objective disordered

Lagerström et al. (2013) quantitative structural objective disordered

Trend: qualitative à quantitative Underrepresented: dynamic, subjective

Visualizing the Hidden Structure of Application Landscapes §  Calculation base: AL topology (applications, information flows) §  Calculation: transitive dependencies of each application

Classification §  Largest cyclic group à Core §  More outgoing dependencies à Control §  More incoming dependencies à Shared §  Less incoming dependencies à Periphery

Propagation cost §  Part of the AL affected by change §  Sum of dependencies / applications2

Classification of applications


Lagerstrom, Robert, Carliss Y. Baldwin, Alan MacCormack, and Stephan Aier. "Visualizing and Measuring Enterprise Application Architecture: An Exploratory Telecom Case." Harvard Business School Working Paper, No. 13-103, June 2013.

2

3 4

5

1

6

7

8 9

Control

Core

Shared Periphery

Complexity of Enterprise Architectures §  Elements (amount & heterogeneity) §  Relationships (amount & heterogeneity)

Calculation of heterogeneity §  Shannon entropy §  No effect of proportional changes §  Significant impact of small changes

Example §  Heterogeneity of database systems

EA complexity metric based on heterogeneity


0

0,2

0,4

0,6

0,8

1

Oracle DB2 SQL Server MySQL

EM = 0.7 EMA = 2 N = 4

Schütz, A.; Widjaja, T.; Kaiser, J. (2013). Complexity in Enterprise Architectures - Conceptualization and Introduction of a Measure from a System Theoretic Perspective. European Conference on Information Systems (ECIS); Utrecht, Netherlands.

Data collection §  6 companies (Financial services and Automotive) §  More than 20 metrics found

Metrics on Application level §  Number of Business Functions (3/6) §  Number of Infrastructure Components (4/6)

Metrics on Domain level §  Number of Applications (4/6) §  Number of Information Flows (6/6) §  Standard conformity (4/6) §  Number of Function Points (3/6) §  Functional redundancy (6/6)

Domain

Reoccurring AL complexity metrics in practice


Application

Application

Application

Semantic processing of legal texts for IT compliance

1.  Interpreting legal texts is non-trivial §  > 6000 laws and regulations in Germany §  Words and expression are hard to understand §  Uncertain, abstract, indeterminate legal terms

§  adequate, effective, appropriate etc. §  International agreements and regulations

2.  Compliance is desirable but expensive

3.  Information systems can support compliance during the §  creation, §  exploration, §  search, §  interpretation and §  visualization processes.


Basel II / III

Sarbanes-Oxley Act

REACH



Company

Assets Objectives Tasks Employees

IT Requirements (Business IT Alignment)

Requirements Engineering

IT Systems

COBIT TOGAF

Controlling

Support through IS Compliance

Requirements (Legal Obligations)

§ Information-

systems LexInform, Juris,

RIS, …

Laws KWG, TMG,

BDSG, …

Authorities (e.g. BaFin)

searching, exploration, interpretation, change tracking etc.



Compliance

Controlling Requirements (Legal Obligations)

§ Information-

systems LexInform, Juris,

RIS, …

Laws/ Regulations KWG, TMG,

BDSG, …

Authorities (e.g. BaFin)

searching, exploration, interpretation, change tracking etc.

§44 IT-examination, auditing, (internal/external) revision, etc.

1.  Information Retrieval (IR)

§  Searching, finding and exploring of information in unstructured documents §  Meet the demand of information

2.  Artificial Intelligence (AI)

§  Automatically derive new information / knowledge §  Answer questions:

§  How has process XY be implemented in order to be compliant? à NO automation but decision-support

Collaborative knowledge work is ubiquitous in organizations


Development of large

software systems

Solving complex problems in communities

Producing new ideas and

innovations

How can software support processes for collaborative knowledge work?

Theoretical basis of the research project involves three different disciplines


Knowledge Work

Literature on knowledge work in organizations provides an understanding of the problem. Description of the problem: •  Characteristics of knowledge

work •  Complex vs. Complicated

problems •  Roles in knowledge work

Adaptive Case Management

Adaptive case management is a novel approach to support knowledge-intensive processes. Solution ideas from ACM: •  Essential requirements for ACM

support •  Emergent design of processes •  Evolution of processes with

templates

Social Principles and Patterns

Knowledge work relies on the successful collaboration of different roles. Facilitating collaboration: •  Building successful online

communities •  Learning from existing

communities on the web •  Principles and patterns

Goal Orientation •  Describe which goals should be achieved •  Goals guide the stream of work •  Replaces traditional process model

Emergence •  Empowerment and participation of end users •  Adaptability of templates at run-time •  Continuous improvement of templates

Data Centricity •  Data as driver for knowledge work •  Goal-oriented transformation of data •  Integration of processes and data

Collaboration •  Knowledge creation through interaction •  Building a successful online community

Case Templates •  Sharing and preservation of knowledge •  Access to recurring best practice patterns

Solution: Empowering users to collaboratively structure knowledge-intensive processes


Create a new task for „Neue Idee“

Logical and temporal dependencies with CMMN

Adding a new task

Attribute types

Drag and drop of attributes on tasks

Access rights on attributes

Completed tasks

Hide completed tasks

Unstructured information

In-place editing

New attribute for the template

Des

ign

Prin

cipl

es

§  Flexible stage-gate process for Innovation Management

§  Development of a future Enterprise Architecture state

§  Artefact-oriented Requirements Engineering processes with templates C

ase

Stu

dies

Analysis of related work and identification of research questions for three domains.

!

!

!

Evaluation 1

Evaluation 2

Evaluation 3

Prototype for collaborative structuring of knowledge-intensive processes.

1. RESSCOPE EARCH

Derivation of requirements for an Adaptive Case Management solution.

2. LITERATURE REVIEW

3. PROTOTYPE

Case studies to support processes for all three investigated domains.

4. CASE STUDIES

Qualitative evaluation of the three case studies with expert interviews.

5. EVALUATION

Deliverable: Transcript of expert interviews

Deliverable: Implemented prototype

Deliverable: Research questions

Deliverable: Requirements for Adaptive Case Management

Deliverable: Prototype applied in three sample domains

?

?

?

EA Management

Innovation Management

Requirements Engineering

Spreadsheets 2.0

Business users love spreadsheets §  Declarative and interactive paradigm to capture functional dependencies §  Modeling, analysis, simulation, visualization §  Empowerment of business-users §  Emergent structures (data, logic)

Limitations of spreadsheets §  Collaborative work §  Complex linked data

social networks, logistic networks, IT architectures, product models, multi-project plans §  Software Engineering Qualities

modularity, reusability, typing, binding, naming


Motivation

Spreadsheets 2.0: Analysis of complex linked data


Hierarchical data structures Networks

Bank

Geschäft

IT

Unternehmens-steuerung

Handel

Kredit

Andere Produkte

Prozesse

Anwendungen

Infrastruktur

Support

Accounting

Controlling

Reporting

Compliance

For more information visit Spreadsheet 2.0 (http://wwwmatthes.in.tum.de)

Visualizations Functions / Transformations Data



𝑓

𝑓

𝑓

𝑓

𝑓

𝑓

𝑓

𝑓

𝑓

𝑓

𝑓

Users For more information visit Spreadsheet 2.0 (http://wwwmatthes.in.tum.de)


System vision §  Hybrid Wiki data model §  Transparency through pipes & filters architecture §  Functional query language (à la LINQ, Scala, …) §  Intuitive interactive web-based user experience §  Fully integrated in collaboration environment §  Optimized „real time“ evaluation Research questions §  User interface concepts and design (data, functions, views)? §  How do users work with historic data and time series? §  Language design (DSL, familiarity ó expressiveness)? §  System architecture and integration with emerging “big data” technologies? §  Evaluation strategies? §  Optimization strategies (materialized views, …)?


For more information visit Spreadsheet 2.0 (http://wwwmatthes.in.tum.de)

Information systems for problem solving

Puzzle


Reproductive Thinking (Heuristics, Algorithms etc.)

Productive Thinking (Creativity etc.)

Wicked Problem Problem

Example

Information System Support

Degree of Automation

Measuring temperature, …

Business Model Generation, …

Sensors, Embedded Systems, Robotics, Databases, …

SAP R/3, Word Processing, Spreadsheet Software, …

Collaborative Informationsystems, e.g. Wikis, Dropbox, …

Problem

Accounting, …

Degree of Collaboration

IS support for a complex problem: Business model generation


•  Re-use benefits of existing tools and methods •  Business Model Canvas

•  Common terminology •  Visual representation

•  Computer-Aided Morphological Analysis •  Basic problem solving process structure •  Interactive model of the problem/solution space •  Clustering of similar business models

•  Multi-user support

•  Group facilitation support •  Alternate between individual and collaborative phases

è avoid social bias

•  Alternate between convergent and divergent phases è promote creativity

•  Alternate between anonymous and identified interactions è avoid social loafing, increase (constructive) social competition

Work-in-progress: currently implementing prototype, designing process model

Colva: Collaborative learning video annotations

Motivation §  Increasing amount of online learning / lecture / teaching / demonstration /

knowledge / … videos §  New players: universities, schools, individuals, non-profit organizations,

businesses, media companies, … §  It is difficult for learners and educators to discover new relevant material for a

given topic §  It is difficult for learners to find the exact location where a particular topic has

been covered §  Increase quality of the learners feedback on the education material and way

of teaching Research questions §  What are the inhibitors of the collaborative learning video annotations? §  How the tool for collaborative learning video annotations effects the behavior of

instructors and learners?

Sebis Research Profile 49 © sebis

Colva: Collaborative learning video annotations


A conceptual framework for describing augmented teaching sessions

Preparation Live teaching session Post-processing

Phases

Actors

Instructor

Learner

Plan timing of teaching session

Prepare teaching material.

Present teaching material

[Take or review notes.]

Activity

Plan timing of teaching session. (verb) (nouns) activity content involved in the activity

[Take or review notes.] (brackets) optional activities

Colva: A collaborative learning video annotations


Possible synchronous and asynchronous collaboration via video annotations

Phases

Preparation Live teaching session

Post-processing

Act

ors

Instructor - View annotation.

View and create

annotation.

Learner -

Create and view annotation.

Create and view annotation.

Colva: Collaborative learning video annotations on the web


Provide a web solution for collecting learners

annotations during the learning session

Implementation stages Stage 1 Stage 2 Stage 3

Synchronize video-recordings with collected real-time user annotations

Test and evaluate different methods for collaboration through video annotations

usage

Current objective Implement concept in viable prototype

Pilot project

For more information contact Klym Shumaiev [email protected]

© sebis

“Wouldn’t it be nice, if you as a Bachelor student at the faculty of informatics at TU Munich could easily create and manage collaborative annotations aligned with video recordings of the lectures?”

Who?

How? What?

Technische Universität München Department of Informatics Chair of Software Engineering for Business Information Systems Boltzmannstraße 3 85748 Garching bei München Tel +49.89.289. Fax +49.89.289.17136 wwwmatthes.in.tum.de

Florian Matthes Prof.Dr.rer.nat.

17132

[email protected]

Thank you for your attention. Questions?

sebis research profile

Presentations & Public Speaking

Transcript of sebis research profile