topic analysis features

[iulusoy]

• optional pre-cleaning of data: filter words, mentions, hashtags, links - text_clean should be modifiable --> keywords
• topic reduction: for this we need an example - maybe at later stage when required?
• topic hierarchical clustering or or: graph output; for similarity threshold: again need example
• longer texts or: or maybe try out summarizing text first