ARTIFICIAL INTELLIGENCE NEWS SERIES: Artificial Intelligence in Education

by İTÜ | Apr 24, 2020
“Artificial Intelligence in Education” is the first of the news series that we have prepared on artificial intelligence, which has started to penetrate our lives with increasing use in many areas and is an important concept of the future imagination. Accordingly, we held a comprehensive interview with Assoc. Dr. Berk Canberk, Manager of ITU Artificial Intelligence and Data Science Application Research, on artificial intelligence and training in the field of increasing human resource needs in this field.

Can you first tell us what artificial intelligence is? What are the study areas?

Artificial intelligence is the realization of the work done by using human intelligence through computers and algorithms. Although the first artificial intelligence studies were made in the 1980s, artificial intelligence was later interrupted for many years. Artificial intelligence studies have become popular again in recent years as the capacity of processors and databases has increased, and most importantly because we have more data available.

In order to be able to do artificial intelligence, basics such as statistics, machine learning, and data mining are used. In addition to these, sciences such as psychology, medicine, economics, linguistics and logic are also required depending on the field of study. The branches of artificial intelligence are machine learning, deep learning, natural language processing, wireless communications, 5G, unmanned aerial vehicles, cyber security, speech, optimization, robotics, image processing.

Artificial intelligence takes place in three stages as learning, reasoning and validation. During the learning phase, data is collected and transformed into information. The algorithm that best achieves the desired result is determined during the reasoning phase. Verification is the improvement of the algorithm. [1]

“Developments have started to occur in many sectors thanks to artificial intelligence.”

In which sectors do we see artificial intelligence mostly?

Rapid advances have been achieved in recent years thanks to artificial intelligence techniques in voice, speech, communications, visual recognition and perception technologies. Basically, thanks to these technologies, there are developments in many vertical sectors.

For example, we can command voice recognition on our mobile phone, we can automatically group the pictures we take and select the pictures we are in, or we can automatically add subtitles to the pictures according to their content. There are different mathematical models and approaches behind such artificial intelligence applications [2].

Expectations from artificial intelligence are high in text editing, simultaneous translation systems, advice systems such as products or personnel, healthcare and cancer diagnosis, banking, oil, telecommunications, computer software and many other areas. Another noteworthy area is to work on using artificial intelligence in judicial systems in the future.

Some world famous artificial intelligence applications; Deepho, the chess application that defeated Kasparov, IBM, Alpho-Go application developed for Go game, Blue Brain Project that conducts research on the human brain, IBM's IBM Watson artificial intelligence application, finance application TurboTax, text editing application Grammarly and Textio are simultaneous translation video conferencing application Webex and iOS personal assistant Siri, driverless tools, face recognition and emotion understanding.

What happens when artificial intelligence and natural intelligence are compared?

Intelligence can generally be defined as “reasoning, problem solving and learning ability”. Besides human intelligence, there are many different abilities such as emotional intelligence and consciousness. However, comparisons can be made between artificial intelligence and human intelligence in terms of processing speeds. While human intelligence can perform 1 million trillion operations per second, a computer system with the size of a basketball court can perform 93 thousand trillion operations per second [3]. Considering the developments in computer systems, it can be thought that artificial intelligence can capture or even pass natural intelligence.

When people are looking for answers to a question, they make a decision using their past experience, considering several possible alternative ways, taking into account the profit or loss they will eventually get. Artificial intelligence can calculate any scenario that human intelligence cannot foresee in a limited time and find the optimum solution among those scenarios.

An interesting approach in the comparison of artificial intelligence and natural intelligence was introduced by Alan Turing in 1950. According to Alan Turing, Turing Test is done to understand that an artificial intelligence application has reached natural intelligence. In this test, a person asks questions to another person who is out of the field of view and to the application of artificial intelligence on the computer screen. In this communication, which takes place only with text, without sound and image, if the person cannot distinguish which one is human and which is artificial intelligence, then the test of artificial intelligence application is considered success. [4]

What would be the positive / negative effects of artificial intelligence technology in the future?

Artificial intelligence will make great contributions to increase the efficiency and production of smart cities, smart cars, smart buildings, smart agriculture and smart livestock and many other vertical sectors by using data gathered from the internet of things (IoT) and 5G technology in the future. Thus, more people will have access to the data they need.

Thanks to its contribution to planning, cleaning, security and transportation of cities, people will be able to live more comfortably, especially in big cities [5].

Thanks to the monitoring of patients in health, the development of diagnosis and treatment methods, the average human life will be extended.

As for the negative effects of artificial intelligence;

In the future, artificial intelligence will cause unemployment in many sectors, however, it will create new employment areas depending on technology literacy.

Especially in life-threatening areas such as defense and attack systems, nuclear weapons, radars, and autonomous vehicle systems, entrusting machines to artificial intelligence will be life-threatening [6].

It is also important to emphasize ethical debates in artificial intelligence research in order to use artificial intelligence for the benefit of humanity. For this purpose, Artificial Intelligence and Ethics Commissions have been established both in companies developing leading artificial intelligence technologies and in the European Union and the United States [2].

ITU Artificial Intelligence and Data Science Application Research Center Studies

Dr. Berk Canberk stated that ITU Artificial Intelligence and Data Science Application Research Center is carrying out activities to increase scientific output and national product production in technology by enhancing research, knowledge and trained human capacity in the field of artificial intelligence to increase our country’s high technology export and international competitive power. Canberk also gave information about the studies of our university in the field of artificial intelligence for the increasing need for qualified human resources.

“ITU Artificial Intelligence and Data Science Application Research Center carries out studies that will combine the strong academic infrastructure of ITU with the strong innovation and entrepreneurship infrastructure of ITU ARI Teknokent, which is the most active and entrepreneurial teknokent in our country. In this context, many seminars and webinars are organized and awareness on artificial intelligence is increased.

At the same time, one of the aims of the center is to open new lessons for the training of qualified human resources that the country needs in artificial intelligence; conducting research and scientific meetings, projects, theses; opening courses to companies and institutions; joining the trained human resources to the sector; to raise awareness in the society through workshops and trainings especially for young people and children.

For these purposes, a cooperation protocol was signed between the Ministry of National Education and ITU on the use and implementation of artificial intelligence in education and increasing the scope of guidance services. As a pilot project by ITU Artificial Intelligence and Data Science Application Research Center researchers, a training curriculum on the concepts of artificial intelligence was organized and applied at ITU Natuk Birkan Secondary School.”

Could you tell us about the curriculum you have prepared for ITU Natuk Birkan Secondary School and the projects you plan to carry out in the field of education in the future?

We continued curriculum development, which started with the question "What is the algorithm?", by involving the basic structures and mechanisms of supervised learning and algorithmic bias and unattended learning. With these courses, the main purpose is to provide students with the concept of artificial intelligence, to develop a project with teamwork, and to develop the problem solving, teamwork, critical thinking, communication and creativity skills of the students who will be the future employees. Thus, students will be able to adapt to the changing labor market more easily. Students’ learning to think analytically, solve problems and become good team members will enable them to compete in the job market even if the nature of the work changes. These skills will also provide students with a strong foundation for independent thinking in their careers to choose throughout their lives.

In the continuation of this study, we have undertaken instructor education which was the first phase. It is among our goals to implement a comprehensive curriculum that students can benefit from. Another work we intend to carry out later will be to provide these trainings with a distance education platform that all secondary school students can benefit from and to open new horizons for their students in artificial intelligence education.

What can be the contributions of artificial intelligence technologies to the field of education?

In the report published by one of the information technology research studies, Gartner in 2017, [7] it is stated that in 2022, one out of every five personnel who take part in non-routine work will use artificial intelligence tools. Although technologies related to artificial intelligence are still in its infancy, artificial intelligence technologies will make it easier for employees to do ordinary office work in the coming years, or in more complex jobs such as using automatic driving vehicles or humanoid robots. It is important for young people of high school age to learn and use artificial intelligence concepts at an early age in order to avoid difficulties in using these technologies and gain competence to develop these technologies in the future.

In addition to the benefits of artificial intelligence and education being included in the curriculum, artificial intelligence technologies can provide many benefits to educators and education models. For example, the analysis of student data and the creation of personalized training methods with the information to be obtained at the end of this analysis, modeling the individual and intra-group interactions of the students with social network analysis models, creating guidance tools for collaborative learning, determining the emotional state and reactions of students and modeling with multimodal data, course-specific, assistant guide robot (chatbot) applications that can answer questions can be used to identify students with learning difficulties at an early stage and to produce special solutions for these students.

In addition, projects can be developed for generating and evaluating artificial intelligence-based automated tests in order to automatically evaluate students and their level of knowledge. System-oriented studies can be done to produce artificial intelligence solutions that examine the education system, its management, processes and can make suggestions for improvement.

Today, an education-teaching process has started with technology as never before. How can artificial intelligence-based technologies be used within the scope of distance / online education applications?

The coronavirus outbreak placed the distance education / meeting system at the center of the agenda of the whole world for the first time. This system, which is used by the universities in the world and in our country, is inevitably used by everyone today. Considering the success and education quality of platforms such as Coursera and Udemy that university students are aware of, we can assume that the new system will be adopted by students.

We can start the contribution of artificial intelligence technologies for distance education / meeting applications firstly by increasing the speed and quality. Artificial intelligence technologies are being used in the improvement of network infrastructures such as 5G, Wi-Fi and new studies are being carried out.

Another important expectation from artificial intelligence is that it makes efficient video conferences. Studies are carried out on the subjects such as the extraction of the noises caused by the participants, even the speeches that are not related to the subject, the identification of the participants with the face recognition systems, the identification of the objects in the image, the speeches made and the ability to search later.

For the post-conference, artificial intelligence technologies can be used to organize and easily understand the outputs.

Thanks to the data accumulated with the spread of distance education / meeting systems, studies will accelerate especially in the field of natural language processing, especially face recognition, facial expressions and emotion understanding. Cisco has 5 goals for using artificial intelligence in video conferences with Webex application. The first one is operations such as joining conferences and adding others with simple voice commands. The second is to translate the sentences in the conference with natural language processing into text, and transfer them to work / course follow-up or personal organizer applications such as; “Ahmet is preparing a slide for the lesson”. Third, detecting terminological words and making automatic searches, taking relevant documents and summarizing the keywords after the meeting. Fourthly, artificial intelligence can act as a participant, take actions and summarize decisions. The fifth is that artificial intelligence can act as a strategic team member and make recommendations to increase profitability. The first of these goals has been realized, and currently the second and third goals are being worked on [8].