++++++++++++++++++
* Removed dependency on `NLTK, <https://github.com/nltk/nltk/>`_ as it already is a `TextBlob <http://textblob.readthedocs.org/en/dev/>`_ dependency
* Temporary workaround for `NLTK Issue 824 <https://github.com/nltk/nltk/issues/824>`_ for tox/Travis-CI
* (update 13/01/2015) `NLTK Issue 824 <https://github.com/nltk/nltk/issues/824>`_ fixed, workaround removed
* Enabled ``pattern`` tagset conversion (``'penn'|'universal'|'stts'``) for ``PatternTagger``
* Added tests for tagset conversion
* Fixed test for Arabic translation example (Google translation has changed)
* Added tests for lemmatizer
* Bugfix: ``PatternAnalyzer`` no longer breaks on subsequent ocurrences of the same ``(word, tag)`` pairs on Python3 see comments to `Pull Request 11 <https://github.com/markuskiller/textblob-de/pull/11>`_
* Bugfix/performance enhancement: Sentiment dictionary in ``PatternAnalyzer`` no longer reloaded for every sentence `Pull Request 11 <https://github.com/markuskiller/textblob-de/pull/11>`_ (thanks Arttii)