Toby Geeson | Data Analyst
Power BI IMDB & Netflix Dashboards
Initally, data on the top 250 movies was scraped from the IMDB website (www.imdb.com), uploaded to, and then gathered by myself from kaggle.com's public datasets. The data was then cleaned and visualised to develop the following report:
​
Raw Data
Column Description
rank - Rank of the movie
name - Name of the movie
year - Release year
rating - Rating of the movie
genre - Genre of the movie
certificate - Certificate of the movie
run_time - Total movie run time
tagline - Tagline of the movie
budget - Budget of the movie
box_office - Total box office collection across the world
casts - All casts of the movie
directors - Director of the movie
writers - Writer of the movie
Data cleaning:
I performed several data-cleaning techniques within PowerBI after sourcing the data. Initially, I identified and handled null values by either removing them or using calculated columns to impute missing values. Additionally, the 'movie time' column was represented as '2h 43m', which required transformation into numerical data by splitting the column and replacing textual values. After converting this into minutes via a measure, I created measures to calculate the average runtime of movies.
​
To further refine the data, I applied filtering and conversion of columns to the correct data types in the power query editor. before adding visuals in PowerBI.
Interactivity
Next
Given I felt inspired to do more, I decided to build another report. Data of movie and TV series lists from Netflix is regularly uploaded to Kaggle.com, found here Netflix Movies and TV Shows (kaggle.com). This contains information regarding the director, cast, duration, date, country and more. As with the above my aim was to consolidate an interactive and visually appealing report into one page of a Power BI report:
Data cleaning
Important steps included changing data types, renaming columns, splitting by delimiter to break up duration into series and minuets, and trimming columns to allow for accurate visualisation. Equally, filters were added, including filtering out inaccurate/null values that could not be imputed with accurate information, finding the top n values (i.e, top countries) and looking at years 2000 to present in the line graph.
Other steps included finding the correct colour palate for Netflix from their logo by taking the Hex code and choosing which visuals I found the most value from.