[dataset] Microposts2013 wrapper

[RicardoUsbeck]

Write a wrapper for the Microposts2013 dataset.
Annotate the license, experiment type and language.
Give provenance.
Update https://github.com/AKSW/gerbil/wiki/Licences-for-datasets

[TortugaAttack]

Just to have it somewhere:

type: microposts
licence: CC BY-NC-SA 3.0
language: en
provenance: http://oak.dcs.shef.ac.uk/msm2013/ie_challenge/MSM2013-CEChallengeFinal.zip

[MichaelRoeder]

Please add it to this page: https://github.com/AKSW/gerbil/wiki/Licences-for-datasets

[TortugaAttack]

done

[sagnik]

I am trying to evaluate the algorithms on micropost data sets on my local machine. The wiki mentions that it expects the data in certain formats, specifically:

Microposts2013

gerbil_data/datasets/microposts2013/goldStandard.tsv
gerbil_data/datasets/microposts2013/testSet.tsv
gerbil_data/datasets/microposts2013/TweetsTrainingSetCH.tsv

Microposts2014

gerbil_data/datasets/microposts2014/Microposts2014-NEEL_challenge_TweetsTestSet.csv
gerbil_data/datasets/microposts2014/Microposts2014-NEEL_challenge_TweetsTrainingSet.csv

Microposts2015

gerbil_data/datasets/microposts2015/dev/NEEL2015-dev-gold_v3.tsv
gerbil_data/datasets/microposts2015/dev/NEEL2015-dev-tweets.tsv
gerbil_data/datasets/microposts2015/test/NEEL2015-test-gold_v2.tsv
gerbil_data/datasets/microposts2015/test/NEEL2015-test-tweets.tsv
gerbil_data/datasets/microposts2015/training/NEEL2015-training-gold_v4.ts
gerbil_data/datasets/microposts2015/training/NEEL2015-training-tweets_v2.tsv

Microposts2016

gerbil_data/datasets/microposts2016/Dev Set/NEEL2016-dev.tsv
gerbil_data/datasets/microposts2016/Dev Set/NEEL2016-dev_neel.gs
gerbil_data/datasets/microposts2016/Test Set/NEEL2016-test.tsv
gerbil_data/datasets/microposts2016/Test Set/NEEL2016-test_neel.gs
gerbil_data/datasets/microposts2016/Training Set/NEEL2016-training.tsv
gerbil_data/datasets/microposts2016/Training Set/NEEL2016-training_neel.gs

I have the original datasets downloaded, if you ever wrote a transformer to convert them to the format gerbil expects in, could you please point that to me? Otherwise, if you have a data model for the micropost data, I can write the transformer myself.

[TortugaAttack]

Actually they should just show up (the microposts datasets ara defined in thr dataset.properties with the given paths.
If they do not show up in the web frontend, please check the logs and may create a new issue ;)
Cheerz

[MichaelRoeder]

@sagnik the files that you have listed there should be the original files. At least that are the files that we got from the microposts organizers. If you have different files, or other problems, please open a new issue since this issue is already closed.

[sagnik]

Please see #206, @TortugaAttack , I added the logs in there as well.