Just last year towards the Valentine’s, I produced a laid-back research of your own condition of Coffees Match Bagel (or CMB) therefore the cliches and you may styles I watched into the on line profiles female typed (released towards the another site). Although not, I didn’t has actually hard affairs to give cerdibility to the thing i spotted, only anecdotal musings and you may preferred terminology I seen when you are searching because of a huge selection of users presented.
First off, I had to find a method to get the text message analysis regarding the mobile app. The new network investigation and local cache was encoded, therefore instead, I got screenshots and went it using OCR to discover the text. Used to do some by hand to find out if it can functions, also it proved helpful, however, going right on through a huge selection of profiles by hand copying text message so you’re able to an Bing piece could be tedious, and so i had to speed up this.
The knowledge out of CMB is actually tilted in favor of the individual’s individual reputation, and so the investigation I mined throughout the users I watched are tilted into my needs and you will cannot represent every pages
Android possess a nice automation API called MonkeyRunner and you may an unbarred resource Python version entitled AndroidViewClient, which allowed full usage of the Python libraries We already got. All this was imported to the a bing sheet, next installed to help you a beneficial Jupyter computer where We went alot more Python programs using Pandas, NTLK, and you can Seaborn to filter out from research and you will generate the latest graphs less than. Continue Reading…