Project proposal

[nw2455]

Group project proposal @JonathanReeve @Abhimanyu210100 @Rishav1999

[mmmelowang]

@QMSS-G5063-2022/teaching_team SHA: 44466d5986bdfd0cb219fcc75e543ea6fe27e28f

[JonathanReeve]

Re: "Heatmaps of all restaurant locations; highest review count; highest star rating restaurants," I'd be curious as to how much more specific you can make this by asking direct questions of your data. For example:

• What food item(s) appear most often in the most glowing reviews? (I.e., the reviews with the highest number of stars)
• What neighborhoods / cities have the worst-rated restaurants?
• What are the restaurants that are the best value? (I.e., highest rating, but lowest cost). This one could be very useful to the public!
• Where are the MIPs? That is, what restaurants start with poor ratings, then acquire excellent ratings? What about the opposite—restaurants that start out with excellent ratings, then devolve.

One thing I'd want to be careful of is comparing average ratings of restaurants which have many different numbers of ratings. That is, you don't want to compare a restaurant which has a 4.5 average star rating over 100 reviews in 6 months, with another restaurant which has a 4.0 average star rating over 1000 reviews in 6 years. I'd want to make sure to truncate the data so that you're comparing the same number of votes in each. Maybe randomly sample, or maybe only look at votes from a given year or range of years.