The budgeting app is designed to empower users to take control of their finances by providing a personalized, adaptive budgeting experience. With a focus on helping users reach their savings goals, the app utilizes machine learning to analyze spending habits and create customized budget plans. Users can set financial goals, and the app intelligently allocates spending categories, continuously adapting the budget to optimize savings and reduce unhealthy spending patterns.
The app also includes an emotional spending tracker, allowing users to log their emotions when making purchases. By analyzing these emotional inputs, the app helps users understand the impact of emotions on their financial behavior, offering visual insights through easy-to-understand graphs that reveal emotional spending trends. This feature encourages positive behavior changes by making users more aware of the emotional drivers behind their spending.
Additionally, this project incorporates a speech-to-text feature using Wit.AI’s API to enable seamless interaction through natural language processing. Users can speak their spending details, such as and the app will process this information, breaking it down into specific amounts and categories. By leveraging Django as the backend framework, the application allows users to input their spending through voice and have it accurately logged and categorized.
Conceptual Design
The design concept for the budgeting app is a web-based application that utilizes a combination of front-end and back-end technologies, along with machine learning components to deliver a personalized user experience. The overall architecture will be divided into three main parts: the user interface, the backend server, and the machine learning module.
Software Architecture
Front-end: The front-end will be a web interface built with HTML, CSS, and JavaScript, utilizing a modern framework like React to create an interactive and user-friendly experience. The front-end will be responsible for user inputs, displaying insights, visualizations, and interacting with the AI chatbot.
• Backend: The backend will be developed using Python with Django as the web framework. This layer will handle data processing, communicate with the database, and serve requests from the front-end. The backend will also integrate with the machine learning module for real-time predictions and recommendations.
• Machine Learning Module: This module will be responsible for analyzing user spending habits and generating personalized recommendations. It will be implemented using Python, leveraging libraries like pandas for data manipulation and scikit-learn (sklearn) for building machine learning models. The ML models will be used to predict spending behaviors, adjust budget allocations, and provide personalized advice. This module will utilize Jupyter Notebook for EDA and testing machine learning models.
Programming Languages and Frameworks
• Python: For backend development, data processing, and machine learning.
• JavaScript: For the front-end, with React framework to create a responsive and engaging user experience.
• HTML/CSS: For structuring and styling the web interface.
• Django: As the web framework for the backend.
Libraries and APIs
• pandas: To manage and manipulate spending data.
• scikit-learn (sklearn): For training machine learning models to analyze spending behavior and generate
recommendations.
• Matplotlib or Plotly: To visualize trends and insights for the emotional spending tracker.
• Django REST Framework: To create APIs that allow the front-end to communicate with the backend.
• Joblib: save trained models to deploy with Django.
• Wit.ai: handles speech to text
Database
• A SQL database like SQLite will be used to store user information, spending data, emotional tags, and model outputs. This will ensure data is well-organized and accessible for analysis.
Background
The proposed budgeting app uses machine learning for adaptive budgeting, emotional spending analysis, and an AI-powered chatbot for personalized guidance. There are apps in the market which features similar budgeting tools but lack adaptive machine learning and emotional analysis. Apps like Mint Tracks spending but lacks adaptive features. The app adjusts budgets automatically, or PocketGuars shows spendable money but lacks emotional insights. This app links emotions to spending and provides a over insights on how emotions affect spending. The app will combine adaptive budgeting, emotional analysis, and AI to offer a unique, personalized financial experience.
Required Resources
This Project will require resources which includes Django for backend development, React for front-end user interface, SQLite for the database, and Jupyter Notebook for exploratory data analysis and machine learning model training. Machine learning libraries like pandas, scikit-learn, and joblib will be used for data manipulation and model deployment. APIs include the Wit.Ai for natural language processing. Additional tools include VS Code or PyCharm for development and GitHub for version control.
Project Abstract
The budgeting app is designed to empower users to take control of their finances by providing a personalized, adaptive budgeting experience. With a focus on helping users reach their savings goals, the app utilizes machine learning to analyze spending habits and create customized budget plans. Users can set financial goals, and the app intelligently allocates spending categories, continuously adapting the budget to optimize savings and reduce unhealthy spending patterns.
The app also includes an emotional spending tracker, allowing users to log their emotions when making purchases. By analyzing these emotional inputs, the app helps users understand the impact of emotions on their financial behavior, offering visual insights through easy-to-understand graphs that reveal emotional spending trends. This feature encourages positive behavior changes by making users more aware of the emotional drivers behind their spending.
Additionally, this project incorporates a speech-to-text feature using Wit.AI’s API to enable seamless interaction through natural language processing. Users can speak their spending details, such as and the app will process this information, breaking it down into specific amounts and categories. By leveraging Django as the backend framework, the application allows users to input their spending through voice and have it accurately logged and categorized.
Conceptual Design The design concept for the budgeting app is a web-based application that utilizes a combination of front-end and back-end technologies, along with machine learning components to deliver a personalized user experience. The overall architecture will be divided into three main parts: the user interface, the backend server, and the machine learning module. Software Architecture
Database • A SQL database like SQLite will be used to store user information, spending data, emotional tags, and model outputs. This will ensure data is well-organized and accessible for analysis.
Background The proposed budgeting app uses machine learning for adaptive budgeting, emotional spending analysis, and an AI-powered chatbot for personalized guidance. There are apps in the market which features similar budgeting tools but lack adaptive machine learning and emotional analysis. Apps like Mint Tracks spending but lacks adaptive features. The app adjusts budgets automatically, or PocketGuars shows spendable money but lacks emotional insights. This app links emotions to spending and provides a over insights on how emotions affect spending. The app will combine adaptive budgeting, emotional analysis, and AI to offer a unique, personalized financial experience. Required Resources This Project will require resources which includes Django for backend development, React for front-end user interface, SQLite for the database, and Jupyter Notebook for exploratory data analysis and machine learning model training. Machine learning libraries like pandas, scikit-learn, and joblib will be used for data manipulation and model deployment. APIs include the Wit.Ai for natural language processing. Additional tools include VS Code or PyCharm for development and GitHub for version control.
Proof of concept https://github.com/stellakjs/MLbudgetapp.git Slide https://tuprd-my.sharepoint.com/:p:/r/personal/tuj97408_temple_edu/_layouts/15/Doc.aspx?sourcedoc=%7BDFD7B92F-E757-4EDB-B535-E637DBF6F9A9%7D&file=Personalized%20Budgeting%20App%20Using%20Machine%20Learning.pptx&action=edit&mobileredirect=true