Add initial analysis notebook and readme

.gitignore

@@ -87,3 +87,5 @@ ENV/
 
 # Rope project settings
 .ropeproject
+
+data/

README.md

@@ -1,2 +1,28 @@
 # KaggleQuoraQuestionSimilarity
 Github repo for kaggle quora question similarity problem
+
+# Data
+The data size for this compoetition is too large to host in GitHub directly.
+All notebooks in this repo expect the following data to be available on your machine, in the following directories:
+
+data/
+ * sample_submission.csv - from the [competition website](https://www.kaggle.com/c/quora-question-pairs/data)
+ * test.csv - from the [competition website](https://www.kaggle.com/c/quora-question-pairs/data)
+ * train.csv - from the [competition website](https://www.kaggle.com/c/quora-question-pairs/data)
+ * wikitext-103/ - from the [wikitext dataset](https://metamind.io/research/the-wikitext-long-term-dependency-language-modeling-dataset/)
+   * wiki.test.tokens
+   * wiki.valid.tokens
+   * wiki.train.tokens
+ * wikitext-2/  - from the [wikitext dataset](https://metamind.io/research/the-wikitext-long-term-dependency-language-modeling-dataset/)
+   * wiki.test.tokens
+   * wiki.train.tokens
+   * wiki.valid.tokens
+ * wikitext-103-raw/ - from the [wikitext dataset](https://metamind.io/research/the-wikitext-long-term-dependency-language-modeling-dataset/)
+   * wiki.test.raw
+   * wiki.valid.raw
+   * wiki.train.raw
+ * wikitext-2-raw/ - from the [wikitext dataset](https://metamind.io/research/the-wikitext-long-term-dependency-language-modeling-dataset/)
+   * wiki.test.raw
+   * wiki.train.raw
+   * wiki.valid.raw
+