BenKlein's post
cancel
Showing results for 
Search instead for 
Did you mean: 
Level 9

Just for Fun - Analyzing my "Review JSON"

Hello Local Guides, 

 

Just for fun, I decided to Takeout My Data

Here are some quick factoids for 528 reviews: 

1) Less than a paragraph: On average I have 75 words per post. (Min 14/Max 475)

2) A start on my writing career: Outside of edits in total, I have 39852 words or 80 pages of my novel. 

3) I am very local (1): 88% of my reviews are in my home country (CAN), with 11% in my neighboring country (USA), and a mere 0.2% outside of these behemoths. 

4) I am very local (2): 77% of my reviews come from my consortium of local municipalities within 45km of my home. 

5) I am positive: 3.8 stars are what my places get, with a total of two 1 star ratings for the places I truly dislike. 

6) What to get there in: 98% of my posts reference the words parking and transit. 

7) I am "Pretty" repetitive: Sadly outside of articles, I use Pretty most often :-(. 

 

Preparing the document (I used Excel because I figure if someone wants to duplicate this they don't want my poorly designed Python): 

 

1) First I downloaded my Googlemaps data (I didn't get all my data just the maps) from Take Out (Link Above) and unzipped it. 

2) Select "Reviews JSON" and uploaded to a JSON conversion site to get it into a CSV (not linking in this article but google it) 

3) Add a column to the right of the heading "properties/Review comments"

4) Add a formula in each cell of this column =IF(LEN(TRIM(A2))=0,0,LEN(TRIM(A2))-LEN(SUBSTITUTE(A2," ",""))+1) to count each cell. 

5) You can now manipulate this data with Min, Max, Average, Sum formulas. If you have a lot of zero's you can use Countif to reduce those issues. 

6) Add several columns (6/7/8?) to the right of the address for each location. 

7) Use the text to columns functionality to split this data by commas (be aware that if you don't want to overwrite you need a lot of columns). Be aware that the data has extra comma's for Malls, major facilities, some odd mapping issues and special places. 

😎 Clean up the data by getting each municipality/city, and country into the same column. 

9) Clean up the data again by removing spaces before municipality/city names. (Search Replace, or VBA code)

10) You can then use countif or vlookup to make some fun correlations. 

11) To get word lookups you can use Excel Functionality but it might be easier to cut and paste that column into Word for manipulation.  (I used multiple word processing and writing assistance programs to review mine).

 

Have fun. 

 

Ben Klein

"Review Responsibly - When in doubt don't review"

 

 

 

Vancouver, BC, Canada
1 comment
Level 7

Re: Just for Fun - Analyzing my "Review JSON"

Neat stuff, @BenKlein!! Just wondering, did you originally do the analysis in Python? Just thinking--it may be cool to extend your data analysis and make some pretty plots and heatmaps or whatnot. Maybe you don't want to show this to the public, but I for one would be really interested in what my reviews look like!