Local Event Sentiment Tracker and Attendance Forecaster

Abstract

The research on “Local Event Sentiment Tracker and Attendance Forecaster” presents a system that helps local event organizers plan better using data instead of guesswork. It focuses on frequently conducted cultural and social events. The system collects comments, reviews and posts from online sources and analyzes whether public opinion about an event is positive, negative or neutral. These sentiment scores, along with past attendance records are used in a regression or machine-learning model to estimate expected attendance for upcoming events. Event details and audience feedback are stored in a database for tracking and analysis. The system also observes how public mood changes over time and converts sentiment results into forecasting features such as polarity score, positive/negative ratio, review volume and sentiment momentum. These features are combined with historical attendance trends and basic event attributes to train a model that predicts turnout. The approach is evaluated using a structured dataset of local events containing labelled comments. Performance is measured using sentiment classification quality (F1 score) and attendance prediction accuracy (Mean Absolute Error, MAE). The expected result is that better sentiment processing produces clearer sentiment trends and improves attendance forecasting compared to simpler methods. The work provides an end-to-end solution that transforms informal feedback into useful sentiment indicators and attendance forecasts, supporting improved planning, marketing decisions, and resource allocation.