hookline

General Entity Annotation Benchmark Framework

gerbil

related project

AGDISTIS

content

Overview of GERBIL
GERBIL is a general Linked Data benchmarking system (formerly used for entity annotation systems based on the BAT-Framework). GERBIL offers an easy-to-use web-based platform for the agile comparison of annotators using multiple datasets and uniform measuring approaches. To add a tool to GERBIL, all the end user has to do is to provide a URL to a REST interface to its tool which abides by a given specification. The integration and benchmarking of the tool against user-specified datasets is then carried out automatically by the GERBIL platform.

If you want to know more, please have a look at our novel paper about GERBIL at the Semantic Web Journal
We also used GERBIL for benchmarking system for question answering

	BAT-Framework	GERBIL 1.0.0	GERBIL1.2.5	Experiment
Wikipedia Miner	✔	✔	(✔)	A2KB
Illinois Wikifier	✔	(✔)	✔	A2KB
Spotlight	✔	✔	✔	A2KB
AIDA	✔	✔	✔	A2KB
TagMe 2	✔	✔	✔	A2KB
NERD-ML		✔	✔	A2KB
KEA		✔	✔	A2KB
WAT		✔	✔	A2KB
Dexter		✔	✔	A2KB
AGDISTIS	(✔)	✔	✔	D2KB
Babelfy		✔	✔	A2KB
FOX			✔	OKE Task 1
FRED			✔	OKE Task 1
FREME			✔	OKE Task 1
entityclassifier.eu			✔	A2KB
CETUS			✔	OKE Task 2
xLisa			✔	A2KB
DoSer			✔	D2KB
PBOH			✔	D2KB
NERFGUN			✔	D2KB
NIF-based Annotator		✔	✔	any

The following table lists the annotators that are currently available and the experiment types they support. Note that some of the A2KB annotators support the D2KB experiment by offering an own API method. Other A2KB annotators can be chosen for a D2KB experiment as well as described in the wiki. However, since the comparison might not be fair, we marked these annotators with (✔) in the table. The same is done for Entity Typing.

	A2KB, C2KB, Entity Recognition	D2KB	Entity Typing	OKE Task 1	OKE Task 2
AIDA	✔	(✔)
AGDISTIS		✔
Babelfy	✔	✔
CETUS					✔
CETUS (FOX)					✔
Dexter	✔	(✔)
entityclassifier.eu	✔	(✔)
FRED	✔	(✔)	(✔)	✔
FREME e-Entity	✔	✔	✔	✔
FOX	✔	(✔)	(✔)	✔
KEA	✔	✔
NERD-ML	✔	(✔)
Spotlight	✔	✔	✔	✔
TagMe 2	✔	(✔)
WAT	✔	✔
xLisa	✔	(✔)
PBoH		✔
NERFGUN		✔
DoSER		✔

The following table lists the datasets that are currently available and the experiment types they support.

	A2KB, C2KB, D2KB, Entity Recognition	Entity Typing	OKE Task 1	OKE Task 2
AIDA/CoNLL-Complete	✔
AIDA/CoNLL-Test A	✔
AIDA/CoNLL-Test B	✔
AIDA/CoNLL-Training	✔
AQUAINT	✔
DBpediaSpotlight	✔
Dercyznski	✔
IITB	✔
KORE50	✔
MSNBC	✔
Microposts 2013-Test	✔
Microposts 2013-Train	✔
Microposts 2014-Test	✔
Microposts 2014-Train	✔
Microposts 2015-Test	✔
Microposts 2015-Train	✔
Microposts 2015-Dev	✔
Microposts 2016-Test	✔
Microposts 2016-Train	✔
Microposts 2016-Dev	✔
N3-RSS-500	✔
N3-Reuters-128	✔
OKE 2015 Task 1 evaluation dataset	✔	✔	✔
OKE 2015 Task 1 example set	✔	✔	✔
OKE 2015 Task 1 gold standard sample	✔	✔	✔
OKE 2015 Task 2 evaluation dataset				✔
OKE 2015 Task 2 example set				✔
OKE 2015 Task 2 gold standard sample				✔
Senseval 2	✔
Senseval 3	✔
UMBC	✔
WSDM 2012	✔

Long term stability

The idea of GERBIL emerged in September 2014 when a couple of articles released at the same time claimed to be state-of-the-art. Especially, those approaches were not easily comparable due to their heterogeneous set-up, dataset use and evaluation metrics. Thus, we decided to build GERBIL and extend the BAT-Framework to break the barriers for people not able to write source code.

GERBIL is now more than 3 years old and has hosted more than 50.000 experiments. It is currently hosted at the research and development unit of the University Leipzig Computation Center and the Paderborn University which keep daily backups to ensure long-term quotability.

The survey data from our paper can be found at GERBIL's GitHub repository.