marketwise_tangerine

Marketwise Round 2 Submission

This repository contains solutions to the four problem statements in IIIT Nagpur's E-Summit'24 Marketwise hackathon. The solutions aim to address various challenges in event management and optimization using data-driven approaches and AI technologies.

Problem Statements

Problem Statement 1: Event Recommendation and Personalization

The goal of this project was to develop an event recommendation system that tailored suggestions based on attendee preferences and past interactions. It involved data analysis, algorithm selection, and real-time customization to enhance attendee satisfaction. The project was divided into three phases:

1. Understanding and Preparing the Data: We conducted exploratory data analysis (EDA), cleaned and preprocessed the data, and engineered relevant features.
2. Crafting a Recommendation System: We built recommendation algorithms.
3. Real-time Customization and Feedback Integration: Real-time recommendation systems were implemented, integrating feedback mechanisms for continuous improvement.

Problem Statement 2: Revolutionizing Event Networking with AI-Powered Lead Capture and Connection Optimization

This project aimed to revolutionize event networking by introducing an AI-powered lead capture system. The system streamlined data collection, facilitated connections between attendees and stall personnel, and optimized the follow-up process. The project comprised three phases:

1. Data Understanding and Preparation: We analyzed the dataset, cleaned and preprocessed it, and engineered features to understand attendee preferences.
2. Development of AI-Powered Lead Capture System: We designed recommendation algorithms and utilized NLP techniques to match attendees with relevant stalls.
3. Optimization and Feedback Integration: Real-time feedback collection mechanisms were implemented to optimize the lead capture system based on attendee satisfaction.

Problem Statement 3: Event Attendance Forecasting

This project focused on developing predictive models to forecast event attendance based on historical data and external factors. The objective was to accurately predict event turnout for better resource planning and event management. The project consisted of three phases:

1. Data Exploration and Preparation: We first generated a dataset and then analyzed, cleaned and preprocessed it, and engineered features related to event attributes and external factors.
2. Attendance Prediction Model Development: We built predictive models using regression techniques and evaluated their performance.
3. Real-time Attendance Monitoring and Adaptation: Real-time attendance monitoring systems were implemented, allowing for model adaptation based on observed attendance data.

Problem Statement 4: Sponsorship ROI Analysis

This project involved analyzing the return on investment (ROI) for event sponsors using sponsorship data and performance metrics. The goal was to quantify the impact of sponsorships and optimize sponsorship strategies for future events. The project was divided into three phases:

1. Data Exploration and Sponsorship Data Collection: We generated a dataset, then analyzed the dataset and cleaned/preprocessed the information.
2. Sponsorship ROI Modeling: We developed models to quantify ROI for event sponsors and evaluated their performance using relevant metrics.
3. Real-time ROI Monitoring and Optimization: Real-time ROI monitoring systems were implemented to track sponsorship performance and adjust strategies based on observed data.

Repository Structure

• PS1: Contains code and documentation for Problem Statement 1.
• PS3: Contains code and documentation for Problem Statement 3.
• PS4: Contains code and documentation for Problem Statement 4.
• Website: Contains the code for the flask app of Problem Statement model deployment as wll as Problem statement 2
• instance: Contains SQLite database.

Contributors

• Parth Singh - Role: Data Scientist
• Ojas Sinha - Role: Machine Learning Engineer