App Board

Dashboard/Stats of apps available on the Apple App Store.

Comment

💡 Inspiration

Being a developer myself, I know that the most challenging part of the development process is neither the development of the application nor the testing of it. Determining what and when to implement is the most troublesome. Hundreds of applications are available on the App Store, both Apple and Android, and not all the developers who made them are prosperous. Imagine if you knew beforehand, in what period, which app would be most thriving. What Genre of application is going to be trending the most? What price to sell your app at to maximize the profits?

🚔 What it does

AppBoard is an analytical dashboard that compiles the insights found in the 736K IOS Apps Dataset. Besides the visualizations available in the dashboard, the Notebooks folder includes all the analysis and visualizations done in the jupyter notebook, along with Data Cleaning and pre-processing.

👷 How I built it

The dashboard follows the UI aesthetic of the bridge web application, and due to the integration of Bootstrap, it is responsive on mobile and smaller devices. Dash and Plotly were used for the implementation of the dashboard.

The first step after obtaining the dataset was pre-processing, achieved using NumPy and pandas python packages. Removal of Null Values and unwanted attributes took place, and the Date-Time format was also corrected.

After pre-processing, analysis and visualization began. Libraries like Seaborn and Matplotlib were utilized for the visualization part.

😥 Challenges I ran into

  1. Initially, I planned to develop a full-stack machine-learning prediction model for a web application embedded in the dashboard. I was not able to complete it due to the limitations of time.

  2. Last year I did a course on developing dashboards using Dash, but it was so long ago that I hardly remembered anything. I had to go through a lot of Plotly documentation to implement the callback feature. Also, the UI part of the dashboard was quite overwhelming as I had to distribute the graphs between the columns impeccably to give it an aesthetic look.

🥇 Accomplishments

With minimal expertise in Dash, Plotly, and Heroku, I created a Live Dashboard after reading documentation, and articles from "towards data science", and watching tutorial videos.

📚 What I learned

  1. Dash and Plotly (Dashboard Development)
  2. Exploratory Data Analysis (Numpy, Pandas, Seaborn, Matplotlib)

💭 What's next for App Board

  1. Integrating Callbacks (Multiple Inputs and Outputs) for a better interactive dashboard.
  2. Adding new Dash Core and HTML Components for a better understanding of the data.
  3. Supplementary Data like application size, update size, etc., for better insights.
  4. Machine Learning Model to predict the price of any app before deploying to the Apple App Store.
Share this project:

Updates