User Device Behaviour Analysis Using Power BI
Project Description:
Understanding how users interact with their devices is critical for improving user experience and optimizing app performance. This report provides insights into user behaviour patterns, such as app usage, battery consumption, and data usage, segmented by device type, operating system, age, and gender. It also categorizes users into behaviour classes, helping identify heavy or light users, which can guide product development and marketing strategies.
Importance of the Project:
The analysis helps companies understand which devices and operating systems are more efficient or popular, allowing them to optimize their apps accordingly. It also aids in identifying user behaviour trends, providing valuable data for targeted marketing, product innovation, and improving customer satisfaction.
Potential Stakeholders:
o?Product Managers: To identify user preferences and optimize app features.
o?Marketing Teams: For targeted campaigns based on user behaviour.
o?App Developers: To optimize app performance on specific devices and operating systems.
o?Data Analysts: To generate actionable insights from user behaviour data.
o?Customer Experience Teams: To enhance user satisfaction by understanding pain points like high battery usage or excessive data consumption.
Project Initialization
o?Project Title: User Device Behaviour Analysis
o?Objective: The goal of this Power BI report is to analyse user behaviour based on device usage, battery consumption, data usage, and categorize users based on their behaviour classes.
o?Key Insights:
o?Identify device usage patterns.
o?Understand how different operating systems influence app usage and battery drain.
o?Analyse behaviour trends by age and gender
Features Descriptions
o?User ID: A unique identifier for each user.
o?Device Model: The model of the user's device.
o?Operating System: The operating system the device runs on (e.g., android, iOS).
o?App Usage Time (min/day): The total time spent using apps per day, in minutes.
o?Screen on Time (hours/day): The time the device screen is on, in hours per day.
o?Battery Drain (mAh/day): The amount of battery consumed per day in milliampere-hour.
o?Number of Apps Installed: The total number of apps installed on the device.
o?Data Usage (MB/day): Data consumed per day, in megabytes.
o?Age: Age of the user.
o?Gender: Gender of the user.
o?User Behaviour Class: A categorical class representing user behaviour.
Data Preparation
o?Dataset: Import the provided CSV file (user_behavior_dataset.csv) into Power BI. (Storage Mode : Import)
o?Data Cleaning:
o?Ensure there are no missing or null values.
o?Convert data types as needed (e.g., Age, App Usage Time, Battery Drain should be numeric).
o Create calculated columns or measures if necessary (e.g., convert App Usage Time (min/day) to hours for consistency).
Data Transformation
o Power Query Editor:
领英推荐
o Load the dataset and clean it if required (e.g., removing duplicates or unnecessary columns).
o?Add calculated columns:
o?Average App Usage per Hour: Create a new column dividing App Usage Time by Screen on Time.
o Data Usage per App: Divide Data Usage (MB/day) by Number of Apps Installed.
Data Model
The dataset has a simple structure with no relationships, so no additional tables or data model adjustments are required unless additional datasets are introduced later.
Report Design
Visuals:
o?User Overview:
o?Card Visual: Display total number of users.
o Pie Chart: Breakdown by Gender.
o Bar Chart: Count of users by Operating System.
o?Device Usage Analysis:
o Column Chart: Display average App Usage Time across different device models.
o Clustered Bar Chart: Show Battery Drain vs. Screen On Time by Device Model or Operating System.
o?Behavioural Insights:
o?Stacked Column Chart: Analyse User Behaviour Class by Age Group and Gender.
o?Scatter Plot: Plot App Usage Time vs. Data Usage to identify high usage patterns.
o?Age and Gender Analysis:
o?Line Chart: Analyse how Battery Drain changes with Age.
o Donut Chart: Distribution of users by Gender.
Adding Interactivity
o Slicers:
o Create slicers for Device Model, Operating System, and Age Group to filter the data dynamically.
o?Drill-through:
o?Enable drill-through on visuals for deeper analysis, such as filtering the User Behaviour Class by Device Model.
Calculated Measures
o Total App Usage Hours: SUM(App Usage Time) / 60.
o?Average Battery Usage per User: AVERAGE(Battery Drain).
o Behaviour Class Proportion: Calculate the percentage of users in each User Behaviour Class.
Final Report Layout
o Page 1: User Overview: High-level KPIs and demographic breakdown.
o Page 2: Device and App Usage: Detailed analysis of device usage and app behaviour.
o?Page 3: Behavioural Insights: Insights into user behaviour class segmentation and patterns by age and gender.
Export and Sharing (Optional)
o Once the report is complete, publish it to the Power BI Service and set up appropriate access permissions for stakeholders.
o Schedule regular data refresh if the dataset is updated periodically.