NLP
in this repository you will find different tools to process text data
facebook crawler
crawls texts from the newsfeed of a specific site
spelling correction
a statistical spelling corrector which uses a dictionary and propability to correct misspellins
wiki_word2vec_domain
creats word embeddings using word2vec on wikipedia dump and allows you to further train the model on domain specific texts
Topic Model (LDA)
uses stopwords and bigrams to reduce dimension extracts topics from text including visualisation
bokeh_cloud
interactiv visualisation of categorial texts
Sentiment RNN
uses a Recurrent neural network to learn and predict entiment from twitter data
Groschenromangenerator RNN
takes german romantic short stories and generates new ones using a Recurrent neural network