Merge pull request #5 from AlexanderSchultheiss/develop

Develop

Author: AlexanderSchultheiss

EXPERIMENTS.md

@@ -278,20 +278,22 @@ Examples of how type and structure information can be transferred to an element-
 
 ### Model Conversion
 Turning a model into an element-property representation requires a pre-processing step which is domain and technology-specific. 
+
+#### EMF Models
 While some of the models used in our experiments were already provided in an element-property representation, the majority 
 of models has been converted from EMF. More specifically, we implemented converters for parts of the UML meta-model 
 implementation shipped with the Eclipse Modeling project, and for customly defined meta-model of architectural models 
 following the component-connector principle.
 
-In order to run the converters on the EMF models provided in the directory [emf-models](emf-models) do the following:
+In order to run the converters on the EMF models provided in the directory [emf-models](models-original/emf-models) do the following:
 * Build the required jar files with Maven: `mvn package`
 * Execute the converters
 ```shell
-java -jar target/UML2CSV-jar-with-dependencies.jar classdiagram emf-models/argouml
-java -jar target/UML2CSV-jar-with-dependencies.jar classdiagram emf-models/bcms
-java -jar target/UML2CSV-jar-with-dependencies.jar classdiagram emf-models/ppu
-java -jar target/UML2CSV-jar-with-dependencies.jar statemachine emf-models/ppu_statem
-java -jar target/Architecture2CSV-jar-with-dependencies.jar emf-models/bcs
+java -jar target/UML2CSV-jar-with-dependencies.jar classdiagram models-original/emf-models/argouml
+java -jar target/UML2CSV-jar-with-dependencies.jar classdiagram models-original/emf-models/bcms
+java -jar target/UML2CSV-jar-with-dependencies.jar classdiagram models-original/emf-models/ppu
+java -jar target/UML2CSV-jar-with-dependencies.jar statemachine models-original/emf-models/ppu_statem
+java -jar target/Architecture2CSV-jar-with-dependencies.jar models-original/emf-models/bcs
 ```
 * The converted models are saved under [`./csv-models`](csv-models)
 
@@ -300,6 +302,9 @@ In a nutshell, we use the generic EMF model traversal and reflective API to acce
 and referenced elements. We exploit domain knowledge (on meta-models) to filter (and condense) some information 
 considered irrelevant (too verbose) for similarity-based matching.
 
+#### Simulink Models
+Information on how to convert the Simulink models can be found in the dedicated [README](models-original/Simulink/README.md).
+
 ### Running the Experiments On Your Dataset
 * Save the dataset's `csv` file to the [`experiment_subjects`](experimental_subjects) directory, next to the other dataset files. (Do not zip your files or locate them in subdirectories. They should be directly at the top-level, e.g., at `experiment_subjects/customdataset.csv`.)
 * Open the experiment properties in a text editor

README.md

@@ -8,7 +8,9 @@
 # RaQuN (Range Queries on N input models)
 
 RaQuN is a scalable n-way model matching algorithm, which uses multi-dimensional search trees for efficiently finding suitable 
-matching candidates through range queries. This repository comprises the artifacts for our paper
+matching candidates through range queries. 
+
+This repository comprises the [artifacts](https://github.com/AlexanderSchultheiss/RaQuN/releases/tag/Camera-Ready) for our paper
 _Scalable N-Way Model Matching Using Multi-Dimensional Search Trees_ which we presented at the 
 International Conference on Model Driven Engineering Languages 2021 
 ([MODELS 2021](https://conf.researchr.org/home/models-2021)).
@@ -19,6 +21,8 @@ Authors:
 [Thomas Thüm](https://www.uni-ulm.de/in/sp/team/thuem/),
 [Timo Kehrer](https://www.informatik.hu-berlin.de/de/forschung/gebiete/mse/mitarb/kehrerti.html).
 
+Additionally, this repository comprises the [artifacts](https://github.com/AlexanderSchultheiss/RaQuN/releases/tag/SoSyM-Submission) for our MODELS'21 SoSyM Special Issue submission.
+
 ### Cite As
 ```bibtex
 @inproceedings{SBG+:MODELS21,

models-original/Simulink/README.md

@@ -0,0 +1,7 @@
+# Simulink Model Preparation
+There are four directories for the Simulink exportation to their element property representation:
+
+- [Simulink_models](Simulink_models) contains the Simulink models, we used in the paper.
+- [element_property_algorithm](element_property_algorithm) contains the Matlab scripts to process Simulink models to their element property representation. For models used by Schlie et al., the scripts also create the different variants from each model. If you want to rerun any of the scripts, you have to adjust the hardcoded directory names to your file system.
+- [input_of_algorithm](input_of_algorithm) contains Simulink models which can be processed by the scripts in "element_property_algorithm". Adapt the hardcoded name in the script to the path of this directory. The models of "Simulink_models" are a subset of the models contained here, but the naming scheme of "input_of_algorithm" works with the scripts of "element_property_algorithm".
+- [output_of_algorithm](output_of_algorithm) contains the element property representations of the Simulink models, we used in the paper. We discarded identical, completely different and trivial sized variants. We renamed the outputs to fit to the ones from "Simulink_models" and our paper's naming scheme.