Remodularizing Legacy Model Transformations with Automatic - - PowerPoint PPT Presentation

KIT – University of the State of Baden-Wuerttemberg and 
 National Research Center of the Helmholtz Association SOFTWARE DESIGN AND QUALITY GROUP 
 INSTITUTE FOR PROGRAM STRUCTURES AND DATA ORGANIZATION, FACULTY OF INFORMATICS

www.kit.edu

Source: pixelio.de

sdq.ipd.kit.edu

Remodularizing Legacy Model Transformations 
 with Automatic Clustering Techniques

Andreas Rentschler, Dominik Werle, Qais Noorshams, Lucia Happe, Ralf Reussner

• 3rd Workshop on the Analysis of Model Transformations

Monday, September 29, 2014

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Model-driven Software Quality Prediction

Performance model


• f a component-based


software architecture Performance data: Execution time Throughput Resource utilisation

Motivation ◉ ⚪ Validation ⚪ Conclusion ⚪ 2 Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Model-driven Software Quality Prediction

Performance model


• f a component-based


software architecture Performance data: Execution time Throughput Resource utilisation

Motivation ◉ ⚪ Validation ⚪

Palladio Component Model Measurement Data Generated Simulation Code SimuCom Simulation Framework Transformation Run

Conclusion ⚪ 2 Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Model-driven Software Quality Prediction

Performance model


• f a component-based


software architecture Performance data: Execution time Throughput Resource utilisation

Motivation ◉ ⚪ Validation ⚪

Palladio Component Model Measurement Data Generated Simulation Code SimuCom Simulation Framework Transformation Run

5k lines of code

Conclusion ⚪ 2 Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Model-driven Software Quality Prediction

Performance model


• f a component-based


software architecture Performance data: Execution time Throughput Resource utilisation

Motivation ◉ ⚪ Validation ⚪

Palladio Component Model Measurement Data Queueing Petri Net Model Simulate Transformation

5k lines of code

QVT-O

Conclusion ⚪ 2 Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

Legacy Transformation

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

Legacy Transformation Modularized Transformation

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

Legacy Transformation Modularized Transformation Manual Decomposition

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

Legacy Transformation Modularized Transformation Manual Decomposition Clusters Cluster Analysis

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

Legacy Transformation Modularized Transformation Manual Decomposition Control and Data Dependencies Dependency Analysis Clusters Cluster Analysis

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

How can we support typical transformation designs? What dependence information is required? Legacy Transformation Modularized Transformation Manual Decomposition Control and Data Dependencies Dependency Analysis Clusters Cluster Analysis

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Problem and Overall Approach

3 Motivation ⚫ ◉ Approach ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

How can we support typical transformation designs? What dependence information is required?

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Design Rules

What’s makes model transformations different from GPL programs? Data-centric operations Data is hierarchically structured Data models extrinsically defined

• Common decompositional styles [Lawley04]:

4 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

Source-driven Target-driven Aspect-driven

• ne-to-many

mappings many-to-one mappings, M2T templates a mixture of both

Approach ◉ ⚪ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

StopAction ActivityModel Source Model Activity Action StartAction actions successors

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

StopAction ActivityModel Source Model Activity Action StartAction actions successors ProcessModel Target Model Process Step steps next

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

StopAction ActivityModel Source Model Activity Action StartAction actions successors ProcessModel Target Model Process Step steps next mapActivity2Process Transformation mapAction2Step call call mapAction2Step mapAction2Step call mapping mapping mapping mapping in

• ut

in in in

• ut

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

StopAction ActivityModel Source Model Activity Action StartAction actions successors CompositeActions Composite Action actions ProcessModel Target Model Process Step steps next mapActivity2Process Transformation mapAction2Step call call mapAction2Step mapAction2Step call mapping mapping mapping mapping in

• ut

in in in

• ut

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

StopAction ActivityModel Source Model Activity Action StartAction actions successors CompositeActions Composite Action actions ProcessModel Target Model Process Step steps next mapAction2Step call mapping call in

• ut

in

• ut

createProcess helper mapActivity2Process Transformation mapAction2Step call call mapAction2Step mapAction2Step call mapping mapping mapping mapping in

• ut

in in in

• ut

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

A Minimalistic Example Transformation

5 Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪

Activity2Process Action2Step CompositeAction2Step StopAction ActivityModel Source Model Activity Action StartAction actions successors CompositeActions Composite Action actions ProcessModel Target Model Process Step steps next mapAction2Step call mapping call in

• ut

in

• ut

createProcess helper mapActivity2Process Transformation mapAction2Step call call mapAction2Step mapAction2Step call mapping mapping mapping mapping in

• ut

in in in

• ut

Approach ⚫ ◉ ⚪ ⚪ ⚪ ⚪ ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

The Bunch Clustering Approach*

6 Motivation ⚫ ⚫ Approach ⚫ ⚫ ◉ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

* by [Mitchell06]; alternative approaches are ARCH, ACDC, LIMBO, … [Shtern12]

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

The Bunch Clustering Approach*

6 Motivation ⚫ ⚫ Approach ⚫ ⚫ ◉ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

* by [Mitchell06]; alternative approaches are ARCH, ACDC, LIMBO, … [Shtern12]

1 1 5 5 15 1 5 5 15 5

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

The Bunch Clustering Approach*

6 Motivation ⚫ ⚫ Approach ⚫ ⚫ ◉ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

* by [Mitchell06]; alternative approaches are ARCH, ACDC, LIMBO, … [Shtern12]

1 1 5 5 15 1 5 5 15 5 1 1 5 5 15 1 5 5 15 5

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

The Bunch Clustering Approach*

6 Motivation ⚫ ⚫ Approach ⚫ ⚫ ◉ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

* by [Mitchell06]; alternative approaches are ARCH, ACDC, LIMBO, … [Shtern12]

Clustering methods: hill climbing, genetic, exhaustive

1 1 5 5 15 1 5 5 15 5 1 1 5 5 15 1 5 5 15 5

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

The Bunch Clustering Approach*

6

clusters i and j: MQ =

k

X

i=1

CFi, CFi = 8 < : 0, µi = 0

µi µi+ 1

2

Pk j=1 j6=i (✏i,j+✏j,i),

• therwise

Motivation ⚫ ⚫ Approach ⚫ ⚫ ◉ ⚪ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

* by [Mitchell06]; alternative approaches are ARCH, ACDC, LIMBO, … [Shtern12]

Modularity quality index: favor low coupling & high cohesion

Clustering methods: hill climbing, genetic, exhaustive

1 1 5 5 15 1 5 5 15 5 1 1 5 5 15 1 5 5 15 5

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Dependence Analysis

7 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ◉ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

mapping_mapAction2Step class_core_Action package_core class_core_BasicAction mapping_mapActivity2Process entry_main class_core_Activity class_core_StartAction mapping_mapAction2Step2 mapping_mapAction2Step3 class_core_StopAction package_process class_process_Step class_process_Process helper_createProcess package_composite mapping_mapAction2Step class_composite_CompositeAction

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Dependence Analysis

7 «package» «package» «package» «write» «write» «write» «read» «read» «read» «read» «read» «read» «call» «call» «call» «call» «call» «call» «write» Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ◉ ⚪ ⚪ ⚪ Validation ⚪ Conclusion ⚪

mapping_mapAction2Step class_core_Action package_core class_core_BasicAction mapping_mapActivity2Process entry_main class_core_Activity class_core_StartAction mapping_mapAction2Step2 mapping_mapAction2Step3 class_core_StopAction package_process class_process_Step class_process_Process helper_createProcess package_composite mapping_mapAction2Step class_composite_CompositeAction

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Weight Configuration

Depending on the style…

• …initialize a configuration vector for weighting the various types of edges:

8 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ◉ ⚪ ⚪ Validation ⚪ Conclusion ⚪

• ntaining package.These weights constitute a part

ector WC := hWwrite, Wread, Wcall, Wpackagei 2 N4

0.

zero naturally results in the respective type of

Method creates/updates model element Method reads model element Method invokes another method Model element is contained in package

Source-driven Target-driven Aspect-driven

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assigning Weights

9

artitions subsume one

• n h1, 15, 5, 15i

for write depen

• ntaining package.These weights constitute

ector WC := hWwrite, Wread, Wcall, Wpackagei 2 zero naturally results in the respective ty ning p := natur

«package» «package» «package» «write» «write» «write» «read» «read» «read» «read» «read» «read» «call» «call» «call» «call» «call» «call» «write» Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ◉ ⚪ Validation ⚪ Conclusion ⚪

mapping_mapAction2Step class_core_Action package_core class_core_BasicAction mapping_mapActivity2Process entry_main class_core_Activity class_core_StartAction mapping_mapAction2Step2 mapping_mapAction2Step3 class_core_StopAction package_process class_process_Step class_process_Process helper_createProcess package_composite mapping_mapAction2Step class_composite_CompositeAction

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assigning Weights

9

artitions subsume one

• n h1, 15, 5, 15i

for write depen

• ntaining package.These weights constitute

ector WC := hWwrite, Wread, Wcall, Wpackagei 2 zero naturally results in the respective ty ning p := natur

«package» «package» «package» «write» «write» «write» «read» «read» «read» «read» «read» «read» «call» «call» «call» «call» «call» «call» «write» Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ◉ ⚪ Validation ⚪ Conclusion ⚪

mapping_mapAction2Step class_core_Action package_core class_core_BasicAction mapping_mapActivity2Process entry_main class_core_Activity class_core_StartAction mapping_mapAction2Step2 mapping_mapAction2Step3 class_core_StopAction package_process class_process_Step class_process_Process helper_createProcess package_composite mapping_mapAction2Step class_composite_CompositeAction

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assigning Weights

9

artitions subsume one

• n h1, 15, 5, 15i

for write depen

• ntaining package.These weights constitute

ector WC := hWwrite, Wread, Wcall, Wpackagei 2 zero naturally results in the respective ty ning p := natur

15 15 15 1 1 1 15 15 15 15 15 15 5 5 5 5 5 5 1 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ◉ ⚪ Validation ⚪ Conclusion ⚪

mapping_mapAction2Step class_core_Action package_core class_core_BasicAction mapping_mapActivity2Process entry_main class_core_Activity class_core_StartAction mapping_mapAction2Step2 mapping_mapAction2Step3 class_core_StopAction package_process class_process_Step class_process_Process helper_createProcess package_composite mapping_mapAction2Step class_composite_CompositeAction

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Example Clustering

10

(SS-L1):mapping_mapActivity2Process (SS-L0):mapping_mapAction2Step

mapping_mapAction2Step class_core_Action

(SS-L0):package_core

package_core class_core_BasicAction

(SS-L0):mapping_mapActivity2Process

mapping_mapActivity2Process entry_main class_core_Activity

(SS-L0):mapping_mapAction2Step2

class_core_StartAction mapping_mapAction2Step2

(SS-L0):mapping_mapAction2Step3

mapping_mapAction2Step3 class_core_StopAction

(SS-L1):helper_createProcess (SS-L0):class_process_Step

package_process class_process_Step

(SS-L0):helper_createProcess

class_process_Process helper_createProcess

(SS-L0):mapping_mapAction2Step

package_composite mapping_mapAction2Step class_composite_CompositeAction

Motivation ⚫ ⚫ Validation ⚪ Conclusion ⚪ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ◉

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assessing Results

Compare derived clustering with a manual expert clustering Using three similarity/distance metrics

11 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ◉ Conclusion ⚪

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assessing Results

Compare derived clustering with a manual expert clustering Using three similarity/distance metrics

11 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ◉ Conclusion ⚪

Table 1: Activity2Process – Manual vs. derived clustering

Configuration Statistics Similarity to expert clustering # C l u s t e r s M Q i n d e x P r e c i s i

• n

R e c a l l E d g e S i m M e C l Expert clustering Derived manually 3 1.067 100% 100% 100 100% Method-call dependencies only Hill Climbing, W C = h0, 0, 1, 0i 2 1.214 20.00% 100% 54.54 60% Class-level dependencies Hill Climbing, W C = h1, 15, 5, 15i 2 1.083 33.33% 100% 72.72 85%

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assessing Results

Compare derived clustering with a manual expert clustering Using three similarity/distance metrics

11 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ◉ Conclusion ⚪

Table 1: Activity2Process – Manual vs. derived clustering

Configuration Statistics Similarity to expert clustering # C l u s t e r s M Q i n d e x P r e c i s i

• n

R e c a l l E d g e S i m M e C l Expert clustering Derived manually 3 1.067 100% 100% 100 100% Method-call dependencies only Hill Climbing, W C = h0, 0, 1, 0i 2 1.214 20.00% 100% 54.54 60% Class-level dependencies Hill Climbing, W C = h1, 15, 5, 15i 2 1.083 33.33% 100% 72.72 85%

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assessing Results

Compare derived clustering with a manual expert clustering Using three similarity/distance metrics

11 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ◉ Conclusion ⚪

Table 1: Activity2Process – Manual vs. derived clustering

Configuration Statistics Similarity to expert clustering # C l u s t e r s M Q i n d e x P r e c i s i

• n

R e c a l l E d g e S i m M e C l Expert clustering Derived manually 3 1.067 100% 100% 100 100% Method-call dependencies only Hill Climbing, W C = h0, 0, 1, 0i 2 1.214 20.00% 100% 54.54 60% Class-level dependencies Hill Climbing, W C = h1, 15, 5, 15i 2 1.083 33.33% 100% 72.72 85%

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assessing Results

Compare derived clustering with a manual expert clustering Using three similarity/distance metrics

11 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ◉ Conclusion ⚪

Table 1: Activity2Process – Manual vs. derived clustering

Configuration Statistics Similarity to expert clustering # C l u s t e r s M Q i n d e x P r e c i s i

• n

R e c a l l E d g e S i m M e C l Expert clustering Derived manually 3 1.067 100% 100% 100 100% Method-call dependencies only Hill Climbing, W C = h0, 0, 1, 0i 2 1.214 20.00% 100% 54.54 60% Class-level dependencies Hill Climbing, W C = h1, 15, 5, 15i 2 1.083 33.33% 100% 72.72 85%

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Assessing Results

Compare derived clustering with a manual expert clustering Using three similarity/distance metrics

11 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ◉ Conclusion ⚪

➡ Significantly better results with model elements considered

Table 1: Activity2Process – Manual vs. derived clustering

Configuration Statistics Similarity to expert clustering # C l u s t e r s M Q i n d e x P r e c i s i

• n

R e c a l l E d g e S i m M e C l Expert clustering Derived manually 3 1.067 100% 100% 100 100% Method-call dependencies only Hill Climbing, W C = h0, 0, 1, 0i 2 1.214 20.00% 100% 54.54 60% Class-level dependencies Hill Climbing, W C = h1, 15, 5, 15i 2 1.083 33.33% 100% 72.72 85%

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Conclusion

Problem: Model transformation programs are often badly structured.

12 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ⚫ Conclusion ◉

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Conclusion

Problem: Model transformation programs are often badly structured.

12 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ⚫ Conclusion ◉

Legacy Transformation Modularized Transformation Manual Decomposition Control and Data Dependencies Dependency Analysis Clusters Cluster Analysis

Approach: Apply automatic clustering algorithms to do the job.

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Conclusion

Problem: Model transformation programs are often badly structured.

12 Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ⚫ Conclusion ◉

Legacy Transformation Modularized Transformation Manual Decomposition Control and Data Dependencies Dependency Analysis Clusters Cluster Analysis

By including the model structure, we can derive clusterings that follow source and target-driven decompositional styles. Selecting ‘good’ weights requires expertise and/or experimenting. Findings: Approach: Apply automatic clustering algorithms to do the job.

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Conclusion

Problem: Model transformation programs are often badly structured.

12

Can we support aspectual decompositions? How do the results compare in maintenance scenarios?

Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ⚫ Conclusion ◉

Legacy Transformation Modularized Transformation Manual Decomposition Control and Data Dependencies Dependency Analysis Clusters Cluster Analysis

By including the model structure, we can derive clusterings that follow source and target-driven decompositional styles. Selecting ‘good’ weights requires expertise and/or experimenting. Open challenges: Findings: Approach: Apply automatic clustering algorithms to do the job.

2014-04-25 Remodularizing Legacy Model Transformations with Automatic Clustering Techniques – Andreas Rentschler et al.

Conclusion

Problem: Model transformation programs are often badly structured.

Thanks!

12

Can we support aspectual decompositions? How do the results compare in maintenance scenarios?

Motivation ⚫ ⚫ Approach ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ ⚫ Validation ⚫ Conclusion ◉

Legacy Transformation Modularized Transformation Manual Decomposition Control and Data Dependencies Dependency Analysis Clusters Cluster Analysis

By including the model structure, we can derive clusterings that follow source and target-driven decompositional styles. Selecting ‘good’ weights requires expertise and/or experimenting. Open challenges: Findings: Approach: Apply automatic clustering algorithms to do the job.