Barni: AI-Powered Proctoring for Academic Integrity
In the rapidly evolving landscape of online education, maintaining academic integrity has become an increasingly complex challenge. As digital learning environments become more prevalent, traditional methods of preventing cheating and plagiarism often fall short. Silicon Valley High School (SVHS) is addressing this issue head-on with the development of Barni, an advanced AI proctoring system designed to protect the validity of online assessments while ensuring a fair and non-intrusive experience for students.
When fully embedded in SVHS’s 60+ courses, this technology will be available to schools adopting the SVHS curriculum, offering a comprehensive solution to the pressing issue of academic dishonesty in digital learning environments.
An Integrated Approach to Academic Integrity
What sets Barni apart from standalone proctoring solutions is its deep integration within the SVHS educational ecosystem. Barni operates seamlessly within the SVHS Smart Editor, which itself is embedded in the broader SVHS Curriculum and Learning Platform. This integrated approach allows for a level of control and security that is difficult to achieve with add-on proctoring tools.
By working within this ecosystem, Barni has access to a wealth of contextual information about the student, the assignment, and the course. This comprehensive view enables more accurate detection of potential cheating attempts while reducing false positives that could unfairly penalize honest students.
Collaboration with Ace: Addressing the Root Causes of Academic Dishonesty
One of the most innovative aspects of the SVHS approach is the collaboration between Barni and Ace, the AI writing assistant. This partnership addresses not just the act of cheating itself, but also some of the underlying causes that drive students to consider academic dishonesty in the first place.
Writer’s block is a common frustration that can lead students to contemplate plagiarism or other forms of cheating. Ace helps alleviate this issue by providing students with guidance on how to start their essays, structure their arguments, and develop their ideas. Importantly, Ace achieves this without generating content or enabling cheating. Instead, it acts as a supportive writing coach, encouraging students to articulate their own thoughts and ideas.
This supportive environment created by Ace can significantly reduce the temptation to cheat. By helping students overcome initial hurdles and maintain progress throughout the writing process, the system aims to build confidence and reduce reliance on dishonest practices.
Creating a Controlled Testing Environment
Barni creates a secure testing space by employing a multi-faceted monitoring approach that closely mirrors the controlled environment of traditional in-person exams. This comprehensive monitoring includes:
Eye tracking through the webcam, which allows Barni to detect suspicious eye movements that might indicate a student is looking at unauthorized resources off-screen. The system can differentiate between normal eye movements and patterns that suggest reading from a hidden source or communicating with someone off-camera.
Keyboard and mouse input analysis is another crucial component. Barni learns each student’s typical typing patterns and can flag sudden changes that might indicate someone else has taken over the test. It can also detect unusual mouse movements or clicks that could suggest accessing hidden resources.
Audio monitoring is employed to detect potential verbal assistance. Barni can identify sounds that might indicate someone is providing answers to the student, while filtering out normal background noise to reduce false alarms.
This multi-layered approach allows Barni to create a secure testing environment that closely replicates the controlled conditions of a physical exam room, but with the added benefits of AI-powered detection and analysis.
Harnessing the Power of Machine Learning
At the heart of Barni’s capabilities is its use of advanced machine learning algorithms. These algorithms are designed to identify patterns indicative of cheating, including subtle behaviors that might escape the notice of human proctors.
The power of this approach lies in its ability to analyze vast amounts of data and learn from each proctoring session. As Barni accumulates more data, its ability to accurately identify suspicious behavior continues to improve. This means the system becomes more effective over time, constantly adapting to new cheating methods as they emerge.
Machine learning allows Barni to detect nuanced patterns that might not be immediately obvious. For example, it might identify correlations between certain types of eye movements, typing patterns, and audio cues that, when combined, strongly suggest cheating behavior. These complex patterns would be extremely difficult for a human proctor to detect consistently.
Furthermore, the system’s ability to learn and adapt is crucial in the ever-evolving landscape of online education. As students develop new methods of cheating, Barni’s algorithms can be updated to recognize these new patterns, ensuring that the system remains effective in the face of changing tactics.
Comprehensive Features for Robust Proctoring
Barni’s feature set is designed to provide comprehensive protection against various forms of academic dishonesty. The system begins with a robust identity verification process that includes facial recognition, voice verification, and typing pattern analysis. This multi-factor approach significantly reduces the risk of impersonation, a common concern in online testing environments.
Once a student’s identity is verified, Barni secures the testing environment by locking down the student’s device. Access is restricted to the SVHS Smart Editor only, preventing the use of unauthorized resources or communication tools. This creates a controlled digital space that mimics the security of a physical testing center.
Throughout the assessment, Barni maintains vigilant oversight. Its continuous monitoring capabilities allow for real-time detection of potential cheating attempts, enabling immediate intervention if necessary. This ongoing surveillance acts as a powerful deterrent, discouraging students from even attempting to cheat.
In addition to monitoring student behavior, Barni works closely with the SVHS Smart Editor to prevent and detect plagiarism. The system performs real-time content analysis, comparing written responses against extensive academic databases. It can identify not just direct copying, but also attempts to paraphrase or reword existing content, providing a comprehensive defense against various forms of plagiarism.
Balancing Security and Student Experience
While Barni is rigorous in its approach to academic integrity, SVHS recognizes the importance of creating a fair and comfortable testing environment for students. The system is designed to strike a careful balance between maintaining security and ensuring a positive user experience.
Multiple data points are analyzed before flagging potential issues, reducing the likelihood of honest students being unfairly accused of cheating. In ambiguous cases, the system can escalate to human review, ensuring fair treatment for all students.
Current & Future Versions
We have early versions of Barni operating with Ace within our Smart Editor, however, implementing Barni into all the 60+ SVHS courses will take some time. The machine learning process involves gathering data from large numbers of students. This will take some time. The first versions of Barni to be deployed in SVHS courses in 2024 do not yet benefit from the advanced machine learning versions that will be introduced in 2025 and beyond.
In Summary
Barni is designed to address the complex challenges of maintaining academic integrity in online learning environments. By combining cutting-edge AI technology with a deep understanding of educational needs, SVHS is creating a system that not only detects and prevents cheating but also supports students in producing their best original work.
As online education continues to grow and evolve, tools like Barni will play a crucial role in maintaining the credibility and value of digital credentials. Through its comprehensive approach to exam security, intelligent plagiarism detection, and commitment to user privacy and experience, Barni is helping to shape the future of fair and honest assessment practices in the digital age.
While no system is infallible, and the landscape of academic integrity will continue to evolve, Barni represents a robust and adaptable solution to many of the current challenges facing online education. As it develops further, it has the potential to significantly enhance trust in online assessments and support the broader acceptance and growth of digital learning environments.
70,863 Enrolled Students