Step-By-Step Configuration1. Download EMBench - current release: embench-ver.1.2.jar
2. Make sure you have Java 1.6 and access to a relational database (e.g., MySQL).
3. Download the raw data you want to use: Baby Names, Amazon, DBLP, DbPedia, IMDB, Lyrics. Alternatively, you can use the shredders for retrieving data from other systems.
Generating DataFirst, create a configuration file that specifies the repository information, the raw data to be loaded, the entities that should be generated, and the modifiers that should be executed (example configuration file: exampleConf.xml). Then execute EMBench using the following command:
java -cp embench-ver.1.2.jar eu.embench.EMBench exampleConf.xml
Evaluation of a Matching AlgorithmFirst, create a class that extends eu.embench.evaluation.EvalMatching. Then, implement methods (i) "setEntities" that converts the embench entities into the entities used by the specific algorithm, and (ii) "compare" that returns true when the algorithm detects a match between the entities. An example class is available here. Then execute EMBench using the following command:
java -cp embench-ver.1.2.jar EvaluateAlgortihm exampleConf.xml
|Last modified: July 2014, Page maintained by: Ekaterini Ioannou, Yannis Velegrakis|