Graph Matching Toolkit

Author: Riesen Kaspar

---

Quick Start

We recommend the following three steps:

STEP 1:

Unzip the archive Sources.zip, integrate the whole framework in an IDE (e.g. Eclipse) and build the project.

STEP 2:

Define a properties file in order to define the parameters of your graph matching task. In the folder properties you find three examples of such properties (for more details on these parameters we refer to our paper:

K. Riesen, S. Emmenegger and H. Bunke. A Novel Software Toolkit for Graph Edit Distance Computation.. In W.G. Kropatsch et al., editors, Proc. 9th Int. Workshop on Graph Based Representations in Pattern Recognition, LNCS 7877, 142–151, 2013.

STEP 3:

Run the graph matching as a java application. The main method is in GraphMatching.java, the sole program argument is an URL pointing to the properties file (e.g.: ./properties/properties_molecules.prop)

---

Change Log

Changes to the original code.

1.1.0 (2017-10-05) b16d7f4

• More concise console output.
• Exception handling: easier debug of issues related to parsing the properties files.
• More details in the result file.
• More comments about how to set the properties in properties_letter.prop.

1.0.1 (2017-06-27) 0b51e0f

• Implemented the cost function for csvDouble for attributes in the form [double, double, ... , double].

1.0.0 (2017-01-30) e833baa

• Version provided to me by Kaspar Riesen.

---

Discalimer

The author of the source code is Kaspar Riesen. Daniele Zambon (dan-zam) is only the owner of the github repository.