top of page

ANALYTICS PROJECTS

RMIT Customer Experience (CX) - Customer Insights Analysis

  • Languages: Python, R

Mining and analysing a high volume of prospective RMIT student web-chat enquiry transcripts to improve how these enquiries are handled, and drive improvements across their digital channels.

 

By using Natural Language Processing (NLP) techniques, such as topic modelling and sentiment analysis, customer unmet needs and pain points were identified, and recommended which department to put more focus/attention towards.

RMIT CX - Topic Modelling - LDA.png
Query Resolved by Enquiry Topic - Local.png
Average Sentiment - Local.png

CO  Emissions & Temperature Anomaly - World Timeline

2

  • Language: R

A Shiny Dashboard that visualises CO  emissions and the change in temperature of many countries around the world from 1880-2017.

The dashboard consists of:

 

- A bubble + choropleth map: With the size of the bubbles in terms of how much a country produces CO  emissions, and the choropleth map for the surface termperature anomalies.

- Time series plots of both temperature anomalies and CO  for each country.

2

2

2

CO2 - Dashboard - Preview.png

Click here to view and interact with the dashboard in a new window.

Average Weekly Value of Internet Sales in the UK - Forecasting

2

  • Language: R

Initial analysis was performed, looking at trend and seasonality of the original times series (January 2007 to December 2019).

 

The best time series model was then determined and used to make forecasts for the next 10 months in the average weekly value of online retail sales in the United Kingdom (UK). 

Avg. Weekly Value Internet Sales - UK Forecast.png
Avg. Weekly Value Internet Sales - UK Residuals.png
bottom of page