AIEMS Project
I designed an AI system that can predict ‘academic dishonesty’ with marginal accuracy
Chapter-1
Project Title: “AIEMS( Artificial Intelligent Education Monitoring System): Development of an Advanced Artificial Intelligent (AI) Monitoring and Feedback System for Online Education ”
Chapter 1
SMART Classroom and AI Technology
As an Electrical Engineering graduate student I enjoyed the challenge of taking different projects that helped to solve real-world problems. With COVID-19 in effect there are several institutions that shifted to online education as the platform offers both convenience and safety during a pandemic. However, in spite of the several advantages of online teaching but the system still lacks a ‘strong’ infrastructure to handle the challenges offered against in person traditional classroom.
It is projected that online education is likely to reach $325 billion in the next 5 years
1.1 The Challenges of in-person and online classroom teaching
1Individualization and its impact on the learning ability has been the focus of interest for a long time [1]. However, in the United States alone; because of the vast diversity in the classroom population, it becomes difficult for an educator to comply with a “standardized technique” during planning their lesson [2]. There are various factors that can account for the effectiveness of the learning processes [3, 4]. Having experience in direct teaching and working with students as a Graduate Teaching Assistant, I realized the spectrum of students and ensuring quality education is a challenge. Questionnaires have been used regularly as a tool to predict any individual’s learning style [5–8]. Learning Analytics, which includes collection, analysis and using of those data [9] has also been suggested to improve the student’s learning experience. However, in most cases, those assessments were particularly used to generalize a classroom populations overall learning pattern rather than utilizing them to compliment any individual student’s learning style.
2Apart from the element of the learning style an important criteria is academic evaluation and ensuring its integrity. According to a study published in Journal of Academic and Business Ethics, it is highlighted that online educators struggle to ensure the integrity of online student’s grade. The challenges include: (i) ensuring the registered student is the one participating in exams and turning in assignments. (ii) individual assignments are not compromised during submission.
Online recording and live proctoring are various methods adopted by most instructors during recent semesters to ensure integrity of online assessments. However, as students get creative in their approach it becomes difficult to monitor effectively. Hence an AI driven proctoring is likely to be more effective and contribute to the overall quality of institutional education.
3 Bolte et al. investigated the effect of emotional state on the ability of intuitive judgements. The study concluded that negative mood could restrict intuitive coherence judgement. The impact of emotional distress due to a pandemic is equally likely to trigger response on learning trajectory and individual performances. However, to the best of our knowledge there is no significant study performed during such crisis or under similar stimulated environment. Understanding of the influence for changing dynamics on individual student performance can allow to effectively help develop an Artificial Intelligent (AI) System that can help improve the quality of online teaching.
1.2 Artificial Intelligence and SMART Classroom
The pandemic COVID-19 has severely affected traditional or regional classroom learning . Such disruptions are more likely to become frequent interruptions with our history of notable pandemics including Black death, Spanish Flu, cholera, Bubonic plague and pandemic Influenza or if a second wave is likely to occur. One of the core delegations during pandemic includes avoiding social gathering which severely impacts the learning and assessment behavior of traditional classroom . Artificial intelligence (AI) can contribute significantly at innovating the current education system during complete or partial transition to online learning. The applied reasoning or cognitive skills and assessment procedures can be utilized at developing an advanced AI system to complement our existing deliberation. During this COVID-19 pandemic, it is probable to trigger emotional responses among student body. This can drastically influence the learning trajectory of individual students.
Artificial Intelligence (AI) techniques has gained wide popularity because of its ability to predict with marginal accuracy and complex problem solving potential .One of the several advantages of using the AI in SMART Classroom setting is its cognitive potential. The algorithm to be used in the development of the AI can eventually help to achieve the goal of personalized learning pace for students with different learning styles and ability. Lo et al. utilized multi-layer feed forward neural network (MLFF) to develop a web-based learning system with focus on student’s cognitive style [10]. Curilem et al. also proposed a mathematical model for an Intelligent Tutoring System (ITS) based on student’s behavior [11].
Apart from the cognitive ability of an AI, the system allows the option of prediction which becomes vital for forecasting. The time series prediction using AI has been used previously in the financial industry [12–14] and also in the medical decision-making process [15]. Similarly, in the education industry, AI had been used for prediction of school dropouts [16] or evading e-learning classes [17].
The concept of utilizing the architecture of AI in the SMART classroom environment will definitely improve the quality of learning in an ideal scenario. The designed architecture is intended towards engineering students attending both traditional and/ e-learning. The system hence requires modification from its initial concept design to suit the needs of a general student population attending either High School or even recently enrolled into college.
1.3 The concept of SMART Classroom
The concept of a SMART classroom that has the ability to utilize both hardware and software components in order to adapt to the necessities of the students [18,19], has been an area of constant research. Aguilar et al. [20] proposed that utilizing an institution’s Learning Analytics in a SMART classroom environment can produce even more efficient results. The application of Wireless Sensors and Internet of Things (IoT) for analyzing social and behavioral patterns in a SMART classroom environment has also been addressed in several testing platforms [21–25]. Hence, with the evolving technology, it is now easier to predict the individual learning profile and therefore analyze and provide for them accordingly. This should eventually improve an individual student academic performance and his motivation to learn.
1.4 Scientific Merit and Significance of the Proposed Conceptual Project
The primary objective is to develop an advanced AI driven online learning system complementing our existing online approach. Artificial Intelligence has already been a topic adopted by most educational institution.
Recently, Microsoft conducted a survey that highlighted the following statistics: 99.4 % agreed that AI is integral for institution’s competitiveness and 92 % agreed to experiment with the technology.
Hence building on improving our current online learning system will improve our competitiveness with other leading institutions. E-learning is the future of education, even after the pandemic is over, it is a good investment both short term and long term. It is projected that online education is likely to reach $325 billion in the next 5 years . Investment in improving this technology is likely to generate streams of potential missed revenue (both domestic and international) since regional barriers is no longer an obligation. According to a report generated by George Mason University and Skidmore College amongst several other factors one of the leading factors that contribute to student-faculty dissatisfaction at online learning includes :
• “Regulation and substantive student-instructor interactivity is a key determinant of quality in online education, leading to improved student satisfaction, learning and outcomes.”
Another concern expressed by professors for online courses include :
• “But when those students take the final exam in calculus or genetics, how will their professors know that the test takers on their distant laptops are doing their own work and not asking Mr. Google for help.”
To summarize, for any educational institute to become a global dominant leader at online education it needs to address the challenges on (i) timely and cost-effective improved feedback mechanism to students (ii) enhanced online proctoring mechanism to ensure academic honesty.
Note: The next chapter focuses on the architecture design and flowchart to understand the model.
References
[1] Smith, Linda H., and Joseph S. Renzulli. “Learning style preferences: A practical approach for classroom teachers.” Theory into practice 23, no. 1 (1984): 44–50.
[2] Yum, Hea-Young. “Learning Style Preferences”.
[3] Sutarto, Joko. “Determinant Factors of The Effectiveness Learning Process and Learning Output of Equivalent Education.” Advance in social science, education and humanities (2017).
[4] Rahman, Roselainy Abdul, John H. Mason, and Yudariah Mohamad Yusof. “Factors affecting students’ change of learning Behavior.” Procedia-Social and Behavioral Sciences 56 (2012): 213–222.
[5] Ortigosa, Alvaro, Pedro Paredes, and Pilar Rodriguez. “AH-questionnaire: An adaptive hierarchical questionnaire for learning styles.” Computers & Education 54, no. 4 (2010): 999–1005.
[6] Tzouveli, Paraskevi, Phivos Mylonas, and Stefanos Kollias. “An intelligent e-learning system based on learner profiling and learning resources adaptation.” Computers & Education 51, no. 1 (2008): 224–238.
[7] Watson, Mark, Michelle McSorley, Cheryl Foxcroft, and Andrea Watson. “Exploring the motivation orientation and learning strategies of first year university learners.” Tertiary Education & Management 10, no. 3 (2004): 193–207.
[8] Petocz, Peter, Anna Reid, Leigh N. Wood, Geoff H. Smith, Glyn Mather, Ansie Harding, Johann Engelbrecht, Ken Houston, Joel Hillel, and Gillian Perrett. “Undergraduate students’ conceptions of mathematics: An international study.” International Journal of Science and Mathematics Education 5, no. 3 (2007): 439–459.
[9] Almonacid, Florencia, Eduardo F. Fernandez, Adel Mellit, and Soteris Kalogirou. “Review of techniques based on artificial neural networks for the electrical characterization of concentrator photovoltaic technology.” Renewable and Sustainable Energy Reviews 75 (2017): 938–953.
[10]Lo, Jia-Jiunn, Ya-Chen Chan, and Shiou-Wen Yeh. “Designing an adaptive web-based learning system based on students’ cognitive styles identified online.” Computers & Education 58, no. 1 (2012): 209–222.
[11] Curilem, S. Glória, Andréa R. Barbosa, and Fernando M. de Azevedo. “Intelligent tutoring systems: Formalization as automata and interface design using neural networks.” Computers & Education 49, no. 3 (2007): 545–561.
[12] Schierholt, Karsten, and Cihan H. Dagli. “Stock market prediction using different neural network classification architectures.” In Computational Intelligence for Financial Engineering, 1996., Proceedings of the IEEE/IAFE 1996 Conference on, pp. 72–78. IEEE, 1996.
[13] Bentz Y, Boone L, Connor J. Modeling stock return sensitivities to economic factors with the Kalman filter and neural networks. In: Proceedings of the Conference on Computational Intelligence in Financial Engineering (CIFEr). New York: IEEE Press, 1996.
[14] Brauner, Erik O., Judith E. Dayhoff, Xiaoyun Sun, and Sharon Hormby. “Neural network training techniques for a gold trading model.” In Computational Intelligence for Financial Engineering (CIFEr), 1997., Proceedings of the IEEE/IAFE 1997, pp. 57–63. IEEE, 1997.
[15] Loch, Tillmann, Ivo Leuschner, Carl Genberg, Klaus Weichert‐Jacobsen, Frank Küppers, Evangelos Yfantis, Michael Evans, Valeri Tsarev, and Michael Stöckle. “Artificial neural network analysis (ANNA) of prostatic transrectal ultrasound.” The Prostate 39, no. 3 (1999): 198–204.
[16] Martinho, Valquiria Ribeiro De Carvalho, Clodoaldo Nunes, and Carlos Roberto Minussi. “An intelligent system for prediction of school dropout risk group in higher education classroom based on artificial neural networks.” In Tools with Artificial Intelligence (ICTAI), 2013 IEEE 25th International Conference on, pp. 159–166. IEEE, 2013.
[17] Lykourentzou, Ioanna, Ioannis Giannoukos, Vassilis Nikolopoulos, George Mpardis, and Vassili Loumos. “Dropout prediction in e-learning courses through the combination of machine learning techniques.” Computers & Education 53, no. 3 (2009): 950–965.
[18] Aguilar, Jose, Priscila Valdiviezo, Jorge Cordero, and Manuel Sánchez. “Conceptual design of a smart classroom based on multiagent systems.” In Proceedings on the International Conference on Artificial Intelligence (ICAI), p. 471. The Steering Committee of The World Congress in Computer Science, Computer Engineering and Applied Computing (WorldComp), 2015.
[19] Aguilar, Jose. “The smart classrooms at the universities as one of the pillars of the e-society.” In eDemocracy & eGovernment (ICEDEG), 2016 Third International Conference on, pp. 138–144. IEEE, 2016.
[20] Aguilar, Jose, and Priscila Valdiviezo-Díaz. “Learning analytic in a smart classroom to improve the eEducation.” In eDemocracy & eGovernment (ICEDEG), 2017 Fourth International Conference on, pp. 32–39. IEEE, 2017.
[21] Winer, Laura R., and Jeremy Cooperstock. “The “intelligent classroom”: Changing teaching and learning with an evolving technological environment.” Computers & Education 38, no. 1 (2002): 253–266.
[22] Stiefelhagen, Rainer, Keni Bernardin, Hazim Kemal Ekenel, J. McDonough, Kai Nickel, Michael Voit, and Matthias Wölfel. “Audio-visual perception of a lecturer in a smart seminar room.” Signal Processing 86, no. 12 (2006): 3518–3533.
[23] Pentland, Alex. “Socially aware, computation and communication.” Computer 38, no. 3 (2005): 33–40.
[24] Madan, Anmol, Ron Caneel, and Alex Sandy Pentland. “GroupMedia: distributed multi-modal interfaces.” In Proceedings of the 6th international conference on Multimodal interfaces, pp. 309–316. ACM, 2004.
[25] Gligorić, Nenad, Ana Uzelac, and Srdjan Krco. “Smart classroom: real-time feedback on lecture quality.” In Pervasive Computing and Communications Workshops (PERCOM Workshops), 2012 IEEE International Conference on, pp. 391–394. IEEE, 2012.
Note: The next chapter focuses on the architecture design and flowchart to understand the model.
