Nowadays, many multilingual environments exist in different domains like academic, professional and research. These environments have made many people struggle with quick information gathering and managing as it is in a different language. This invention will be a Web-Based Intelligent Note Summarizer with Multi-Language Translation, designed to address the challenges of information overload and language barriers in academic and professional settings. The system will utilize advanced Natural Language Processing (NLP) AI models to accurately condense lengthy texts into concise summaries. It will then integrate a sophisticated machine translation service to convert these summaries, or the original text, into the user's preferred language, initially focusing on Malay, English, and Chinese. Beyond core summarization and translation, the platform will feature intelligent functions like automatic highlighting of key concepts and on-demand annotation extraction for definitions. A cloud-based database will allow users to securely save and manage their processed notes. The application will be developed as a responsive web platform, ensuring wide accessibility across devices and browsers, and will incorporate offline functionality for uninterrupted use. This integrated solution aims to significantly enhance productivity, break down language barriers, and facilitate better learning and comprehension for users in multilingual environments.

Invention Technical Description

The system will be architected on a modern, multi-layered stack. The frontend will be built with Next.js and React.js to create a dynamic, single-page application (SPA) that provides a responsive user interface for all interactions. The backend will leverage Google Firebase as a comprehensive Backend-as-a-Service (BaaS) solution. Firebase Authentication will manage user security, while Firebase Cloud Functions will serve as a secure, serverless intermediary to handle the application's core logic. These functions will make API calls to external AI services: an NLP model from Hugging Face Transformers for abstractive summarization, highlight detection, and annotation extraction, and the DeepL API for high-quality translation. Data persistence will be managed by Firebase Firestore, a NoSQL cloud database, for storing user profiles, saved notes, and annotations. For offline capability, the client-side will use IndexedDB to queue user actions when internet connectivity is lost, with a synchronization handler automatically pushing this data to the cloud upon reconnection. All communications will be encrypted, and sensitive API keys will be securely stored environment variables on the server side.

Demostration/ Presentation Video

Poster/ Broucher/ Invention Photo