Abstract

The objective was to make a comparative study of the IoT based Ensemble Predictive system with real-life teacher predictions on student observation during E-Learning. Data is collected from 46 faculties for 188 periods through an opinion-based survey using a questionnaire. Similarly, for the 188 periods the data was collected from the created IoT based Ensemble Predictive System. The system is designed in such a way that it uses five variables, namely: Level of Interaction, No. of Questions Raised, No. of Students in the Class, No. of Concepts Taught in a Period, and Responsiveness of the Students to Questions during Class Hours, to perform the student observation analysis. From observation, it is interpreted that there is no major difference in the solution provided by the faculties and output generated by the IoT Ensemble Predictive system. Further, it was found that there is a remarkable relationship between the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system. But for the item (No. of Students in the Class), there is no significant relationship between the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system. Therefore, it can be interpreted that when the number of students increased or decreased beyond ideal conditions, there is the possibility of deviation in the opinion provided by the faculties and output generated by the system. Also, the calculated R-value is 0.788, meaning there is a 78.8% strong positive relationship between the opinion provided by the faculties and output generated by the system. The estimated R-Square Value (0.621) indicates that it is possible to forecast an IoT based Ensemble Predictive system in the real world with 62.1% efficiency. Furthermore, the evaluated.

Coefficients’ significance value is not more than 0.05 and this confirms that an IoT based Ensemble Predictive system can be forecasted with real-world information and vice-versa.

Keywords: IoT, education system, E-learning

10.1 Introduction

The World Wide Web was very efficient in terms of posting and retrieving data in the beginning stages. The World Wide Web was only a collection of static web pages with links to other sites and users could access the information they needed by surfing and moving from one page to the next [1]. After some time, the situation changed and web 2.0 was born. A user of a website could do significantly more with it than they could with a web 1.0 site, such as reading material and navigating through one page to another page through hyperlinks. The normal person’s engagement with online resources is easily available. Social networking sites like Facebook, YouTube, and Flickr, among others, are examples of web 2.0 applications. We are currently living in the age of semantic web. Computers in web 3.0 are capable of intelligently evaluating information available on web sites, as well as developing and propagating the information they contain [2].

Until recently, the internet was designed to bring people together, enabling them to communicate, exchange information, send messages, and hold virtual conferences. We now have machine to machine interactions in addition to human to human and human to machine interactions thanks to the introduction of web 3.0. This is a huge step forward. System-to-system communication can be shown through a communication linking a temperature sensor and AC [3]. Consider the following scenario: a temperature sensor monitors the room temperature and when it rises beyond a certain threshold, it sends an indication to the AC prompting it to immediately turn on the switch. If the temperature goes down a specific point [4], the temperature sensor will send a second signal to the air conditioner, this time instructing it to turn off the machine. M2M is a technology that is used in the IOT. With the growth of technology, it is now feasible for people to engage with and contribute information on a website in a conventional manner. The information that can be utilized in the reception and transmission of analyzed data may be used to its full capacity by the specific kind of connection of different things that can be achieved via the internet [5].

As examples, a lot of web 2.0 apps are available. YouTube, Flickr, Facebook, and a slew of other sites are among them. Consumers are increasingly familiar with web 3.0, which is a semantic web that enables them to explore with more sophistication. Because of the existence of web 3.0, persons who use computers can understand the availability of information in a more intellectual manner [6]. Web 3.0 can both send and develop information depending on the data it receives from its data sources. The internet may be considered one of the available resources for individuals to interact and communicate with one another, as well as for the spread and acquisition of relevant information. It may also be used for information transfer and in the sophisticated version, video conferencing can be utilized to make information accessible. Apart from the advancement of man-to-man communication, the promotion of web 3.0 may also be beneficial to the advancement of man-to-machine communication. Web 3.0 also allows for machine-to-machine communications [7]. Every day, more smart items that are tailored to certain situations are produced and they are becoming increasingly common in many fields of education, including higher education. Bright education is the most significant component in establishing IoT bright cities that is made possible by virtual learning and digital transformation when it comes to smart services that are backed by developing technologies such as the IoT. Applying IoT in online learning, when utilized in bright cities, allows e-citizens to be connected and imaginative with a higher degree of involvement and teamwork in the learning activity as well as other decisions made by their peers and the community [8].

The implementation of IoT in smart cities has a great impact on online learning processes by offering a comprehensive electronic approach to the online educational community’s resources through a centrally integrated system in a smart fashion. Academic work shows the ability and effect of the IoT in smart city e-education and e-learning processes are expected, measurable, and cannot be overlooked. Hyper-connection linking items, higher degree of access, scalability, and integration of communication networks (RFID and WSNs) are Internet of Things features that might improve the effectiveness of e-learning techniques in “smart settings,” including brighter cities. It can be considered as the most essential tool for supporting the Internet of Things-based learning system, which is paving the way for more efficient ways in the online-education of smart cities in the not-too-distant future, despite its existing challenges. Future research has to stay focused on present online learning methods and their similarity with smart cities, as well as the creation of new approaches based on this fact.

10.2 Review of Literature

Professionals who want to further their careers and students who want to acquire new skills while maintaining their existing work status, as well as individuals who want to obtain training or prepare for certification exams, are increasingly turning to online education. Institutions are finding it increasingly challenging to keep careful monitoring and evaluation of their courses as the number of persons enrolling in online courses grows [9]. However, owing to a shortage of supervisors and tutors, educational institutions are investing extensively in online teaching and learning to allow their students to utilize the network as a platform. High-bandwidth applications like animations, video conferencing, and simulations, all of which may be provided over the Internet, are available to a worldwide audience of networked learners through the Internet [10]. The website offers a collaborative environment for sharing materials with others, as well as online course courses. Sensors are utilized in the Internet of Things (IoT) to collect data from learners when they interact with virtual platforms like (MOOCs) and the data is then transferred to a central database system to be analyzed. The goal of this investigation is to see whether the IoT can be used to improve online learning and teaching. Several data-mining technologies will be deployed inside the central database systems to filter, organize, integrate, and analyze data in order to provide reports for various management levels [11].

The word “e-learning” refers to the application of technology in the field of education and learning. Learners may utilize virtual classrooms, which are given via e-learning, to expand their knowledge without needing to employ traditional learning methods. As a consequence of the fast proliferation of information and communication technology in today’s environment, the letter “e” has become the symbol of the information technology age we are currently living in [12]. The letter “e” is a frequent abbreviation for electrical components. As seen by the emergence of phrases like “e-learning,” “e-health,” “e-business,” “e-government,” and many others, the prefix “e” is becoming increasingly popular in every industry. As a consequence of globalization, networking, and information technology reaching their pinnacle, e-learning is becoming increasingly significant in the field of education [13]. Today’s world is controlled by globalization, networking, and IT. In eLearning, social media plays an essential role in enhancing the learner experience. In recent years, the use of social media to promote course material, training programs, learning tools, and registration in new courses has grown in popularity. The most extensively used social media sites for marketing include YouTube, LinkedIn, Twitter, Facebook, and Google Plus. The below table gives you a detailed number of online social platforms in India. IoT is a widely utilized technology that allows communication and collaboration through the Internet. It grows in both size and dimension as it develops, affecting many aspects of our life, including educational chances [14]. According to the Internet of Things’ basic premise, any devices having a unique IP address are allowed to connect with one another in the future, both physically and digitally. The data acquired by sensors, tags, or actuators and transferred to a cloud system through a gateway is the most basic structure of the IoT. In the IoT, machine-to-machine with object-to-machine interactions are prevalent. The IoT has facilitated the expansion of a wide range of applications, from simple domestic appliances to complex medical equipment [15]. The Internet of Things covers a broad variety of human activities, including smart cities, smart businesses, and smart energy use to name a few.

Education is one of the most important activities of humans that the Internet of Things has influenced, with the shape of education in the not-too-distant future likely to be converted into a more imaginative framework. IoT is a relatively new phenomenon that promotes innovation across a broad variety of businesses. The area of education is one of these sectors (sometimes known as e-education). IoT has the potential to deliver a broad variety of e-educational technologies that will have a major influence on the future of educational institutions since it can be combined with other information technology (IT) technologies. Intelligent technologies will be installed in the future educational facility [16]. Students and teachers must verify their position as users by showing their finger prints and RFID ID cards in front of the reader, as well as mobile checking, whether visiting the rooms in person or gaining permission to the school’s automated system administration. In the future, sensors will be deployed in IoT classrooms to ensure that only authorized teachers and students have access to the space. The RFID or WSN devices that will be mounted on the smart whiteboards and desks will be able to detect and identify the persons who use them physically. Students and teachers in the smart classroom may connect with one another in a mutually advantageous manner. The Internet of Things (IoT) has the potential to provide secure interaction, linking all physical and technological objects. As a consequence of this technology, students may connect online to labs, libraries, didactic materials, tests, and instructional messages while also doing administrative tasks more efficiently in a large virtual classroom. Another benefit of this kind of virtual learning is that in this format, all learning assignments and activities will be described as objects. E-learning also refers to the incorporation of electronic tools, software and hardware applications, and web-based activities into a learning system or system of learning, and it dates back to the 1980s. In fact, due to the fast development of technology and communication technologies, it is now feasible to study online using a large learning environment, which is becoming more popular (Internet) [17].

The Internet of Things (IoT) allows learning environments to be expanded by integrating actual and virtual things without interfering with the learning process. Traditional e-learning has the ability to provide learners with a wide virtual access environment as a digital strategy, but it also has important limitations, as noted below [18]. The most major constraints in e-learning settings are geographical location, face-to-face communication between objects, and effective cooperation between virtual and real agents. The usage of smart objects in the learning environment is one potential answer to the issues listed above. The Internet of Things (IoT) is commonly recognized as the most important source of smart agents for e-learning. The Internet of Things (IoT) has the ability to include two key aspects into traditional e-learning: intelligence and object interaction [19]. IoT has the potential to establish a vast platform for students and instructors to access a variety of distant learning devices and products. A high degree of interaction between virtual and physical goods may lead to the creation of a wide range of collaborative situations.

10.3 Methodology

The descriptive and experimental research design is adopted for further study. The data is collected from 46 faculties for 188 periods through an opinion-based survey using a questionnaire. Similarly, the 188 periods the data was collected from created an IoT based Ensemble Predictive System. The system is designed in such a way that it uses five variables, namely: Level of Interaction, No. of Questions Raised, No. of Students in the Class, No. of Concepts Taught in a Period, and Responsiveness of the Students to Questions during Class Hours, to perform the student observing analysis.

10.4 Analysis and Interpretation

Analysis is done to find out the average score provided by the system and opinions provided by the faculties during 188 hours.

From the Table 10.1 Group statistics - teacher & system feedback, mean score calculated using the opinion provided by the faculties and output of the analysis made by the system, it can be interpreted that students have a high level of observation during the class hours, as the mean score lies between (4.3404 - 4.4255).

Analysis is done to check the difference in opinion between the faculties and system.

From Table 10.2 Independent sample test - teacher & system feedback, the approximated value is more than 0.05, meaning the null hypothesis is accepted. Therefore, there is no major difference in the answers provided by the faculties and output generated by the IoT Ensemble Predictive system.

Therefore, analysis is taken forward to identify whether there is any significant relationship between the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system.

From Table 10.3 Correlation analysis - teacher & system feedback, the approximate value is less than 0.05 for most of the items, meaning the null hypothesis is rejected. Therefore, there is a closed relationship between the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system.

But for the item (No. of Students in the Class), the evaluated value is greater than 0.05 for most of the items, meaning the null hypothesis is accepted. But, there is no serious relationship between the answers provided by the faculties and output generated by the IoT Ensemble Predictive system when the No. of students increased or decreased beyond normal ideal conditions.

Herein, an analysis was carried out to identify the forecasting efficiency of the IoT based Ensemble Predictive system with the real world.

From Table 10.4 Regression analysis - teacher & system feedback, the Calculated R-value is 0.788, meaning there is a 78.8% strong positive relationship between the opinion provided by the faculties and output generated by the system. The estimated R-Square Value (0.621) indicates that it is possible to forecast an IoT based Ensemble Predictive system with in the real world with 62.1% efficiency. Also, the ANOVA significance value is less than 0.05, illustrating the model is fit. Furthermore, the estimated coefficients’ significance value is less than 0.05 and this confirms that an IoT based Ensemble Predictive system can be forecasted with real-world information and vice-versa.

10.5 Findings and Conclusion

From the evaluation, it can be interpreted that there is no big difference in the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system. Further, it was identified that there is a significant relationship between the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system. But for the item (No. of Students in the Class), there is no significant relationship between the opinion provided by the faculties and output generated by the IoT Ensemble Predictive system. Therefore, it can be interpreted that when the No. of students increased or decreases beyond normal ideal conditions, there is the possibility of deviation in the opinion provided by the faculties and output generated by the system. Also, The Calculated R-value is 0.788, meaning there is a 78.8% strong positive relationship between the opinion provided by the faculties and output generated by the system. The estimated R-Square Value (0.621) indicates that it is possible to forecast an IoT based Ensemble Predictive system in the real world with 62.1% efficiency. Furthermore, the evaluated coefficients significance value is less than 0.05 and this confirms that IoT based Ensemble Predictive system can be forecasted with real-world information and vice-versa.

References

1. 1. Berecz, “Overview of E-learning Strategies from the Point of View of Higher Education,” J. Appl. Multimed., vol. 13, no. 4, pp. 117–127, 2019, doi: 10.26648/jam.2018.4.003.
2. 2. Hammi, R. Khatoun, S. Zeadally, A. Fayad, and L. Khoukhi, “IoT technologies for smart cities,” IET Networks, vol. 7, no. 1, pp. 1–13, 2018, doi: 10.1049/iet-net.2017.0163.
3. 3. M. Njeru, M. S. Omar, S. Yi, S. Paracha, and M. Wannous, “Using IoT technology to improve online education through data mining,” Proc. 2017 IEEE Int. Conf. Appl. Syst. Innov. Appl. Syst. Innov. Mod. Technol. ICASI 2017, no. May, pp. 515–518, 2017, doi: 10.1109/ICASI.2017.7988469.
4. 4. S. Kusuma and D. Kasi Viswanath, “IOT and Big data analytics in e-Learning: A technological perspective and review,” Int. J. Eng. Technol., vol. 7, no. 1, pp. 164– 167, 2018, doi: 10.14419/ijet.v7i1.8.11540.
5. 5. M. Bayani, K. Leiton, and M. Loaiza, “Internet of things (IOT) Advantages on e-learning in the smart cities IoT-based library automation & monitoring system view project Internet of things (IoT) advantages on e-learning in the smart cities,” Int. J. Dev. Res., vol. 07, no. January, pp. 17747–17753, 2017, [Online]. Available: http://www.journalijdr.com.
6. 6. S. Madakam, R. Ramaswamy, and S. Tripathi, “Internet of Things (IoT): A Literature Review,” J. Comput. Commun., vol. 03, no. 05, pp. 164–173, 2015, doi: 10.4236/jcc.2015.35021.
7. 7. Y. S. Sai and K. K. Kumar, “Internet of things and its applications,” Int. J. Eng. Technol., vol. 7, pp. 422–427, 2018, doi: 10.14419/ijet.v7i2.7.10758.
8. 8. V. D. Soni, “IOT connected with e-learning,” Int. J. Integr. Educ., vol. 2, no. 5, pp. 273–277, 2019.
9. 9. K. Balakrishna, F. Mohammed, C. R. Ullas, C. M. Hema, and S. K. Sonakshi, “Application of IOT and machine learning in crop protection against animal intrusion,” Glob. Transitions Proc., vol. 2, no. 2, pp. 169–174, 2021, doi: 10.1016/j.gltp.2021.08.061.
10. 10. Abdellatif et al., “Communication-Efficient Hierarchical Federated Learning for IoT Heterogeneous Systems with Imbalanced Data,” Futur. Gener. Comput. Syst., vol. 128, pp. 406–419, 2021, doi: 10.1016/j.future.2021.10.016.
11. 11. K. F. Muteba, K. Djouani, and T. O. Olwal, “Deep reinforcement learning based resource allocation for narrowband cognitive radio-Iot systems,” Procedia Comput. Sci., vol. 175, no. 2019, pp. 315–324, 2020, doi: 10.1016/j. procs.2020.07.046.
12. 12. J. Su, S. He, and Y. Wu, “Features Selection and Prediction for IoT Attacks,” High- Confidence Comput., p. 100047, 2021, doi: 10.1016/j.hcc.2021.100047.
13. 13. E. Anthi, L. Williams, A. Javed, and P. Burnap, “Hardening machine learning denial of service (DoS) defences against adversarial attacks in IoT smart home networks,” Comput. Secur., vol. 108, p. 102352, 2021, doi: 10.1016/j. cose.2021.102352.
14. 14. M. W. Rahman, R. Islam, A. Hasan, N. I. Bithi, M. M. Hasan, and M. M. Rahman, “Intelligent waste management system using deep learning with IoT,” J. King Saud Univ. - Comput. Inf. Sci., no. xxxx, 2020, doi: 10.1016/j. jksuci.2020.08.016.
15. 15. B. Song et al., “ScienceDirect IoT and Machine learning for in-situ process control using Laser Based IoT and Machine learning for in-situ process using Laser Based Additive study Additive Manufacturing (LBAM) case study new methodology and physical architecture of a , ^⋆ analyze the functional existing products for an assembly oriented Tiwari a product family identification,” Procedia CIRP, vol. 104, pp. 1813– 1818, 2017, doi: 10.1016/j. procir.2021.11.306.
16. 16. D. Deng, X. Li, V. Menon, M. J. Piran, H. Chen, and M. A. Jan, “Learningbased joint UAV trajectory and power allocation optimization for secure IoT networks,” Digit. Commun. Networks, no. August 2020, 2021, doi: 10.1016/j. dcan.2021.07.007.
17. 17. R. Akhter and S. A. Sofi, “Precision agriculture using IoT data analytics and machine learning,” J. King Saud Univ. - Comput. Inf. Sci., no. xxxx, 2021, doi: 10.1016/j.jksuci.2021.05.013.
18. 18. V. R. Kebande, R. A. Ikuesan, N. M. Karie, S. Alawadi, K. K. R. Choo, and A. Al-Dhaqm, “Quantifying the need for supervised machine learning in conducting live forensic analysis of emergent configurations (ECO) in IoT environments,” Forensic Sci. Int. Reports, vol. 2, no. July, p. 100122, 2020, doi: 10.1016/j.fsir.2020.100122.
19. 19. M. Esmail Karar, A.-H. Abdel-Aty, F. Algarni, M. Fadzil Hassan, M. A. Abdou, and O. Reyad, “Smart IoT-based system for detecting RPW larvae in date palms using mixed depthwise convolutional networks,” Alexandria Eng. J., 2021, doi: 10.1016/j.aej.2021.10.050.