Presentations
Presentation files available for download, providing an overview of the project in a concise manner.
Inception
Elaboration
Construction
Transition
Identify synthetic text with ease
using our customizable browser extension.
AI Text Detector is a software system that offers a browser extension to detect synthetic text in web pages and PDF files. It uses sophisticated language model-based metrics to evaluate the probability of text being synthetic.
Our AI text recognition Chrome extension provides accurate identification of human-generated and AI-generated text, enhancing the reading experience for users. The extension's ease-of-use and reliable results make it an essential tool for researchers, students, and professionals alike.
Our AI text recognition extension comes with comprehensive documentation, ensuring users can seamlessly integrate it into their workflow. From installation to advanced customization, our documentation provides all the necessary information to maximize the extension's capabilities.
Customization is key, and our AI text recognition extension offers users the ability to personalize the highlighting color and other settings, making it a versatile tool. This feature is user-friendly, enhancing the user experience and empowering users to tailor the extension to their preferences.
Our AI text recognition extension is designed to be extensible, allowing developers to integrate their own metrics and models. This feature ensures the extension can be adapted to meet the needs of different industries, adding to its versatility and functionality.
Our sleek and intuitive UI/UX design is based on extensive research and user feedback, delivering an easy-to-use interface. This ensures that the AI text recognition extension can be used effortlessly by users, enabling them to make informed decisions about the text they're reading.
Our AI text recognition extension supports both web pages and PDFs, providing consistent results regardless of content type. The extension's versatility makes it a powerful tool for researchers, academics, and professionals who require accurate text recognition in both web and document formats.
We used OpenUP, an iterative and agile software development process, which consists of four phases: Inception, Elaboration, Construction, and Transition. In each phase, we had to achieve a set of milestones before moving on to the next phase.
During the Inception phase, the team identified the problem of the increasing use of synthetic text generated by language models and the need for a tool to detect it. The project's objective was to develop a browser extension that could identify synthetic text and highlight it in different colors based on the probability of it being synthetic.
During the Elaboration phase, the team developed a stable architecture plan for a browser extension that could effectively detect synthetic text. We identified similar projects, refined our requirements, and designed some UI mock-ups. The skeleton structure included actors and use cases to represent responsible entities and functionality needed for the project's success.
During the Construction phase, the team diligently executed the development plan outlined during Elaboration. We followed coding standards, and used appropriate technologies to build a reliable and scalable solution. Rigorous testing and debugging ensured the stability of the implemented functionalities. Integrating user feedback, we made iterative improvements to enhance the overall user experience.
During the Transition phase, we deployed the backend and Model Hub web platform on a server provided by IEETA, ensuring a reliable hosting environment. Additionally, we made the solution easily accessible to users by releasing an extension on the Chrome Web Store. This allowed for a smooth transition from development to deployment.
Presentation files available for download, providing an overview of the project in a concise manner.
Inception
Elaboration
Construction
Transition
Report and other documents available for download, documenting and detailing the work flow of the project.
Report 30/05/23
Abstract 30/05/23
Promotional video to promote the product.
Poster for Students@DETI to summarize information concisely and generate discussion.
Architecture of our AI text detection Chrome extension, designed for seamless integration.
Project management calendar for streamlined team coordination and progress tracking.
| Week | Task | Contributors | Status |
|---|---|---|---|
| 21-02-2023 | Project planning | Alexandre, Daniel, João, Ricardo | ✅ |
| 28-02-2023 | Requirement gathering | João, Ricardo | ✅ |
| 28-02-2023 | Low-fidelity prototype | Daniel | ✅ |
| 28-02-2023 | Report - State-of-art and Architecture | Alexandre | ✅ |
| 07-03-2023 | Actors and Use cases | Ricardo | ✅ |
| 07-03-2023 | Architecture design | Alexandre | ✅ |
| 14-03-2023 | Extension setup and framework testing | Daniel | ✅ |
| 14-03-2023 | Json request elicitation | Alexandre | ✅ |
| 14-03-2023 | API setup and deployment for testing | Ricardo | ✅ |
| 14-03-2023 | Language model probing | Alexandre | ✅ |
| 14-03-2023 | Popup and global button - Initial development | Daniel | ✅ |
| 14-03-2023 | LM Hub - Initial development | João | ✅ |
| 14-03-2023 | Web scraping and Rest API - Initial development | Ricardo | ✅ |
| 21-03-2023 | Extension development and API data display | Daniel | ✅ |
| 21-03-2023 | LM Hub development | João | ✅ |
| 21-03-2023 | Functional LMs and semi-automatic testing for LMs | Alexandre | ✅ |
| 28-03-2023 | Authentication | João | ✅ |
| 28-03-2023 | Database setup | Alexandre | ✅ |
| 28-03-2023 | Integration of scraping function in the extension | Daniel, Ricardo | ✅ |
| 28-03-2023 | Extension UI development and settings | Daniel | ✅ |
| 28-03-2023 | Improve Scraping - Client side | Ricardo | ✅ |
| 28-03-2023 | PDF support - Initial development | Ricardo | ✅ |
| 04-04-2023 | Analyse selected text, LM selection and cancel/redo scan | Daniel | ✅ |
| 04-04-2023 | Blackbox for submitted LMs | Alexandre | ✅ |
| 04-04-2023 | API endpoints to get and upload LMs | Ricardo | ✅ |
| 04-04-2023 | LM processing in backend and storing in database | Alexandre, Ricardo | ✅ |
| 04-04-2023 | LM Hub frontend remake | Ricardo | ✅ |
| 02-05-2023 | Report - first sections | Daniel | ✅ |
| 09-05-2023 | Fix minor bugs in scraping | Ricardo | ✅ |
| 09-05-2023 | Complete PDF support (functional PDF viewer) | Ricardo | ✅ |
| 09-05-2023 | Automate SQL and django connection | Alexandre | ✅ |
| 09-05-2023 | Report | Daniel | ✅ |
| 09-05-2023 | Paste and scan feature | Daniel | ✅ |
| 16-05-2023 | Poster | João | ✅ |
| 16-05-2023 | Cache mechanism to speed up redo scans | Daniel | ✅ |
| 16-05-2023 | Video | Ricardo | ✅ |
| 16-05-2023 | Extension Shortcuts | Daniel | ✅ |
| 23-05-2023 | Model Evaluator | Alexandre | ✅ |
| 30-05-2023 | Core system deployment | Alexandre | ✅ |
| 30-05-2023 | Web extension deployment | Daniel | 🟨 |
Meet our dedicated team of students and mentors from Universidade de Aveiro.
