rbc_google_reviews_data_analysis(rbc Challenges)

dash board look like this

Inspiration

price
learning is fun ## What it does The project is a dashboard build using kibana, the data visualization on the dashboard is to will help RBC employee to better understand the customer reviews and then to improve themself. ## How I built it
data collection using google place API. My limitation on data collection is that because I do not have a business google account, I can only get 5 reviews from one RBC branch location. To collect more data, data from various location is collected
put collected raw data into elasticsearch, this is to save the data for future use, so we don't need to recollected every time using it.
process the raw data by removing non-text data. removing stop word, removing words that shows too frequently or too rare.
build a LDA machine learning model using python gensim to group the data into different topic. This is to analysis the reason that customer give a good review or bad review in the later steps
put the processed data into different index in elasticsearch for visualization purpose
create a data visualization dashboard of the processed data

Challenges I ran into

The main challenge I run into is data collection. In the beginning I wanted to used twitter API, but I have to apply to have a developer account of twitter in order to use that. Then I switched to google API, I realized that I can only have 5 reviews from one location because I don't have a business account. The data would be more useful If I can get all reviews from branch.