Need a generic topic miner and sentiment mining code/app in R/Python that takes a CSV file, reads a particular column in it that contains a multiple sentence text, cleans it (stopword removal, punctuation, case, synonyms, etc), breaks down into tokens, creates n-grams and categorizes each verbatim basis the n-grams into broad categories/clusters. The categorization should be done using multiple classifiers / clustering algorithms and there should be flexibility to choose the one that performs best and generate relevant visualizations on top of it (word-cloud, classification accuracy, etc). The work should preferably be done in Python and detailed code with documentation / comments should be shared. The code should almost be plug and play and should work with sample files that will be shared.
This is a test project that will be followed by a more complex text mining solution with more specifics around tuning the app.