added data/sts/semeval-sts/2017

data/sts/semeval-sts/2017/LICENSE.txt

@@ -0,0 +1,17 @@
+The evaluation sets for STS 2017 are distributed under a Creative Commons Attribution-ShareAlike 4.0 International License:
+
+https://creativecommons.org/licenses/by-sa/4.0/
+
+-----------------------------------------------------
+The Stanford Natural Language Inference (SNLI) Corpus
+-----------------------------------------------------
+STS Tracks 1, 2, 3, 4a, 5 and 6 are based on data drawn from the SNLI corpus.
+
+The Stanford Natural Language Inference Corpus by The Stanford NLP Group is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. Based on a work at http://shannon.cs.illinois.edu/DenotationGraph/.
+
+https://creativecommons.org/licenses/by-sa/4.0/
+
+-------------------
+WMT News Commentary 
+-------------------
+STS Track 4b is based on data drawn from the WMT News Commentary corpus from the annual Conference on Machine Translation (WMT), formally the Workshop on Statistical Machine Translation. 

data/sts/semeval-sts/2017/README.txt

@@ -0,0 +1,54 @@
+ SemEval 2017 Shared Task 1
+ Semantic Textual Similarity (STS)
+
+This package contains the test sets for the 2017 Semantic Textual Similarity
+(STS) shared task. Each evaluation set has the following tab-separated format:
+
+ * One STS pair per line.
+ * Each line contains the following fields: STS Sent1, STS Sent2
+
+The input files are provided in UTF-8.
+
+Example:
+
+Un perro salta sobre los obstáculos para un show canino. A dog jumps over something.
+
+One file is provided for each track of STS 2017, with the exception of track 4.
+Track 4 is subdivided into two subtracks, one for data drawn from SNLI and
+another from data sourced from WMT news data. All other datasets were sourced from
+SNLI.
+
+Input Files (UTF-8):
+--------------------
+STS.input.track1.ar-ar.txt
+STS.input.track2.ar-en.txt
+STS.input.track3.es-es.txt
+STS.input.track4a.es-en.txt
+STS.input.track4b.es-en.txt
+STS.input.track5.en-en.txt
+STS.input.track6.tr-en.txt
+
+Output Files: 
+-------------
+For each evaluation set, please generate a plain text output file with one line
+for each STS pair that provides the score assigned by your system as a floating
+point number:
+
+0.1
+4.9
+3.5
+2.0
+5.1
+
+Release History
+---------------
+v1.0 Jan 16, 2017 - Initial release of the test set
+v1.1 Jan 18, 2017 - Correction to track 1, Arabic-Arabic STS 
+
+Output files should be uploaded to the STS 2017 CodaLab site:
+
+https://competitions.codalab.org/competitions/16051
+
+Each team can submit up to three runs. 
+
+Good Luck!