so due to low workload these days, i was getting really bored, when one of my friend posted a data mining model about forecast of corona outbreak....i got interested in data mining and needed a project.....so i decided to web-scrape n text mine this thread to answer some of my questions using technology - P.S: I did this just for fun, one may need some research for interpreting wordcloud/comparisoncloud although it's very simple for a curious mind, there are much powerful models available in market which i have'nt used yet , I am yet to clean the data only for FR and remove all the "spam" post, which i will be doing in few days if time permits .
Few questions which i wanted to answer were: -
Exploratory analysis : most discussed bars, most discussed monger, sentiment analysis (term taken from web) to predict which bars are above avg.,most famous coatwala etc.
Descriptive analysis : give scoring to bars based on FR (something which basu1970 did based on his exp. ), bar with most +ive FR, most talked about chick (will be difficult as most don't name the chick in FR), approx. damages of every bar (again subjective to monger, n we already have classification of SB<DB<OB but still just for fun), , alias name in of mongers with multiple ID's based on writing style., forecast bar performance based on type of review we give here (though to be completely valid will require access to bar-manager records of no. of footsteps)
It will be interesting to know what more questions do anyone want to get answered while reading the forum and venturing out for hunt in bars?......