This project presents a mobile-based solution called MyPOI, designed to analyze sentiment from reviews of Malaysian Points of Interest (POIs). By integrating Named Entity Recognition (NER) and Deepseek-R1, the system improves both accuracy and efficiency in verifying feedback. Built with React Native and Firebase, MyPOI empowers POI managers with deeper insights for better decision-making and smarter tourism strategies.

Invention Technical Description

This project presents a Mobile-Based Sentiment Analysis System for assessing Malaysian Places of Interest, a mobile application developed to assess the sentiment of tourist's reviews for Malaysian Places of Interest (POIs). The system utilizes Natural Language Processing (NLP) techniques, including sentiment analysis and Named Entity Recognition (NER) to categorize reviews into aspects such as environment, price, and family-friendliness. It integrates Deepseek-R1 for advanced contextual understanding and fact-checking, significantly improving the accuracy and reliability of sentiment review assessments. Built with React Native for cross-platform compatibility (iOS and Android) and supported by Firebase for backend services and real-time data management, the system features an intuitive dashboard that visualizes sentiment scores, a conversational chatbot for on-demand fact verification, and a dynamically ranked list of POIs based on aggregated user sentiment. Designed with accessibility in mind, the application serves both tourists seeking reliable information and POI managers aiming to enhance service quality through data-driven insights. This prototype represents a scalable, AI-enhanced solution to misinformation in tourism information, supporting more informed decision-making and promoting Malaysia’s cultural and natural attractions.

Demostration/ Presentation Video

Poster/ Broucher/ Invention Photo