Finnish educational system committed to gradually introduce AI in education by means of «Elements of AI» and «Building AI» designed by the University of Helsinki. To date, than 750.000 students have attained these courses.
The use of the Artificial Intelligence (AI) in education is progressively evident by the organising of learning and assessment methods. However, to ensure its absolute implementation in education, there is still a long way to go. This paper assesses those particularly students difficulties which pose a challenge to educational institutions. To this day, it is a fact which has not been solved. We are talking about basic necessities such as cognitive abilities for special education needs children, failure at primary school or lack of motivation in secondary education. This study aims reflecting on the need for progress in instruction and learning and the solutions that AI offers us. Genre, tense and number educational reforms are not enought. “Education is an ornament in prosperity and a shelter in adversity”, according to the prestigious scientist, philosopher and polymath, Aristotle.
Key words: #AI #ArtificialIntelligence #educationalsystems #evaluationsystems #educationalstages
An education system is a set of institutions and elements that set the order and function from the different educational stages (preschool, primary, secondary and higher) regulating the function of education in a country [1], according to Isabel Valenzuela, director in intermediate education in Liceo Beltran Prieto. It was also important to maintain a focus on the different learning strategies in an educational system. Nisbet and Shucksmith (1986) see them as test sequences from procediments and activities selected for facilitating the acquisition, storage, and the use of information and knowledge [2].
We are now in the time on information technology in which the younger children are surrounded by multiple devices. Maybe these devices are still unknown for them, but they will follow in their work in the future. From the moment to which each generation is more intuitively and connected to, their role has become apparent. The younger children are capable to understand how some technological devices works, for example, an electronical book. However, technology has also dangers, such as the case of the girls from Michigan, who bought toys via Alexa [3] or the 10-years-old girl, who entered a currency into a plug when Alexa possed her a challenge [4].
In this table it is observed an ascendent evolution in reference to the age in which children have their own smartphone. The source of reference is Ofcom, British communication medium [5].
Age | Percentage of children who have smartphone |
9 years old | 23% |
10 years old | 50% |
15 years old | 94% |
Over the years, there has also been a high development involved in the interaction with intelligent devices from the child population; Alexa and Google Assistant [6]. The source of reference is Ofcom, British communication medium.
Year | Percentage about the interaction with Alexa and Google Assistant |
2018 | 15% |
2019 | 27% |
Several initiatives have flourished in recent years by linking the educational sector with the Artificial Intelligence one. The main objetive has been to improve the education, while facilitating the learning of students by creating personalized materials, and doing tasks that require time and as evaluation [7]. One of the more recent proposals related to Artificial Intelligence in the education is The Recommendation on the Ethics of Artificial Intelligence by UNESCO (2021). This project proposes to incoporate appropriate knowledge with regard to Artificial Intelligence at every level of education [8].
The evaluation of educational systems is controlled by the OECD’s Programme for International Student Assessment. It is a regular analysis carried out every three years by OECD. This analysis measures the academic performance of students in maths, sciences and reading at global level [9].
The last PISA report published in 2018 [10]. Building on the agreed academic performance by OECD and according to this report and the title of this section, there will be represented a number of statistic graphics, that include all the european countries in descending order. On the basis of these figures, we represent the most valued educational systems in EU. To conclude, it will made a final evaluation of the results by mentioning the importance of knowing what the term quality makes reference to in education and how the Programme for International Student Assessment interpret it.
Estonia
In 2002, Estonia included teaching in computational linguistics in the University of Tartu. It is a new oppotunity to study computational linguistics throught special courses from various branches of knowledge: linguistics, maths and computing. To achieve this, students who are taught subjects in higher studies of science (maths, basic knowledge of networks, software and hardware) and AI, they have optional subjects to choose, for instance: theoretical computer science, software systems, and language technologies. As in the case of science, in linguistics higher studies, apart form conventional subjects (syntax, morphology and semantics), the students have the option to choose Artificial Intelligence I & II, morphology and computational lexicology. This project aims to get accuracy in morphology, syntax, and Estonian in language semantics, including a lexical and semantic database. Moreover, from a pragmatic perspective, it is possible to model oral language. The results said it all:
The development of a syntactic-analyzer that generates a spelling checker.
A morphological desambiguator to identify the grammatical categorization, derivations and the origin of words.
A linguistic corpus of the Estonian oral language with more than 300.000 translated words.
Other advantages to improve the educational system of Estonia is the contribution of the University of Tartu in several international projects such as GLOSSER, MULTEXT-EAST, TELRI-I, TELRI-II, CONCEDE Y BABEL, among others [11].
In 2012, Estonia launched the programme «Proge Tiger» managed by the Communication Technologies Foundation for Education and it was founded by the Ministry of Education and Researching of Estonia. The project proposes the introduction of programming and robotics in the national study plans in childhood, primary and professional education [12]. The typical applications of this work are:
In childhood education, teachers use LEGO WeDo, Kodu Game Lab to encourage the learning of students.
In primary school, teachers use Kodu Game Lab, Logo MSW, Scratch, LEGO Mindstorms EV3, programmes, authoring environment of mobile applications, and other applications to teach some subjects (music, maths, physics, biolog), as well as electronics laboratory.
In secondary education and vocational training, professors make use of different programming languages (Python and JavaScript), Codecademy.com courses, 3D graphics, robotics, programmes to have plays, web pages, and applications [13].
In Estonia, the educational institutions have the option to work on programming in the classes to apply in different areas of the curriculum. If the education institutions so decide, they could receive financial compensation to get technological equipment, formation, educational materials, and other necessary resources. The results of this study reveal that the reception on part of the professors responded en masse since from the beginning the project, more than 80 percent of the educational centres of Estonia entered this programme [14].
Likewise, Estonia develops a computational approach to assess the success of studies in an educational institution of higher education. To that end, some methods based on data mining and machine learning are used to identify factors which combate school absenteeism and contribute to evaluate the dropt-out rates. In the end, the level of risk is differentiated by colours; red (high), orange (medium), and green (low). To this classification is added a brief description which says, for example, a student belongs to school drop-outs (red level) in that he hasn’t got enough experience to attend higher education abroad [15].
Finland
The finnish educational system gradually introduces AI in education with programmes such as Elements of AI [16] and Building AI [17]. These online courses were developed by the University of Helsinki. With Elements of AI and Building AI, it is possible to connect all people without requiring specific skills on the topic. It is a way to contribute to technological innovation and adquiring basic notions about Artificial Intelligence [18]. To date, these courses register more than 750.000 students and a global extension of the programme, as they have been coursed in over 170 countries.
Poland
In 2018, Polish universities incorporated the technology of AI in education. One of the initiatives was chatbot, as it can help in the development of technical and programming skills to students, providing the opportunity of adquiring linguistic experience. Simultaneously, students understand the real functioning of AI systems by having into account that AI should always respect civil and human rights. The studied cases in this project have good results because of the promoting lifelong learning and the communicative process without losing sight the aim of educational interaction [19].
Netherlands
In Netherlands, it is developed a flexible approach to introducingly presented AI by applying a new system implemented by Java. It is a course organized by a set of tasks which requires that students use intelligent agents and other techniques based on AI to monitor, filter, and recover important information on Internet. It is a way to engage students in their learning, while achieving intellectual rigour. After the implementing method, we know that students learn more when the traditional approach combines with the constructivist one than using a learning approach to teach programming to students [20].
There are several initiatives to introduce Artificial Intelligence in European education. Particularly, some countries work in learning platforms such as: Coursera, Educalab and edX. These platforms processe a big amount of data, which are generated by the students interactions with learning environments by combining different techniques based on AI [21]. In Spain, for instance, the University Carlos III of Madrid [22], the Technical Univeristy of Valencia [23] implement this initiative. Furthermore, there are other countries: France [24], Germany [25], and Italy [26]. Together with this proposal, other territories offer the option to study AI as a professional career, some of them are: Romania [27], France [28], Poland [29] and Sweden [30].
Following the analysis, it is undoubtedly deny that AI supports the quality of education, because it generates a lot of benefits, in particular: the interaction and encouragement of students and the quick detection of errors in the evaluation process. Despite of being revealing data, we cannot determine, at least in the short run, the total impact of AI in education.
The quality of an educational system is highly complex to determine because it is an ambiguous term, since it implies several concepts: the result approach, the development of aims and magnificence. Moreover, each territorie is suitable to stablish the standard points to get a quality education system [31]. To evaluate the quality of reading, maths and science skills, PISA inform is confined into some criteria [32]:
Reading | Maths | Science |
|
|
|
According to Esteban Sánchez Manzano, associate professor of the Faculty of Education in the Complutense University of Madrid, special education is a set of scientific knowledge and educational, psychological, pedagogical, social and physician interventions intended to improve the potential of people with disabilities [33].
The chapter 1 of the Organic Law 3/2020 regulates that schoolchildren with educational support is one who requires an educative attention outside the ordinary one because of special educational needs and other educational requirements for maximum development of their personal capabilities and, if anything, to meet the overall educational goals for the rest of students [34].
The LOMLOE law establishes that in the primary education is guaranteed educational inclusion, although it is not established as a principle (Art 19). Moreover, the perception is that activating mechanisms to avoid repeating grades. As well, it is also introduced in the article the reduction of ratios student/unit as a measure of attention to special educational needs (Art 20.bis) [35].
The Organic Law on Education (LOE) of 2006 [36] and the Organic Law to improve quality education (LOMCE) of 2013 that modifies it [37], schooling percentage of students which present special educational needs is regulated by principles of standardisation and inclusion by introducing measures of relaxation at the different stages when necessary may be provided. The educational provision into special schools takes place if the needs can not be initiated within the framework of the needs of diversity measures into the ordinary centres.
The four basic principles in special education are:
The normalization means that, as far as possible, a special education student should have the equal rights and obligations that the rest of pupils.
The individualization responds to a professional and therapist intervention; accomodation plan and special methodology.
The compartmentalization is referred to special educational services included in the location where a pupil with a disability lives and develops himself.
The integration comes from the principle of normalization ensuring that these children receive the necessary assistance within the normal groups and non in a disaggregated way [38].
To know what special education needs required the students from different educational stages in the course 2018-2019, we collect the following data in some graphics:
The first makes reference to childhood education. It shows an increase in the number of children who required special attention:
The following graphics is referred to primary education stage. As educational level progresse, we can observe an increase in the number of students with educational needs support. Exceptionally, it has been observed a deceleration in the third grade with serious personality disorders, intelligentsia and pervasive developmental disorders:
This image shows what happen in secondary education. In contrast to childhood and primary education, the number of students with special care is reduced with educational level progresse:
In non-compulsory education, in which students reach majority, the outstanding needs are the same that in the previous educational stages, as distinct from pervasive developmental disorders. And there is no uniformity in these periods.
The results show a higher percentage of students who needed special adaptations. Perhaps, these could be solved by using Artificial Intelligence. In fact, some existing methods could be implemented on special education:
«Google´s Neural Machine Translation System» (GNTS). It is a translation system created by Google in 2016. It uses deep learning and supports more than 100 languages. GNTS directly translates from one language to another without involving English language as interlanguage. This system can be productive for special educational needs, particularly for those children with auditory impairment, as they may make use of the subtitles. In order not to create confusion to students, the captions should be correctly translated because the video understanding and the accurate transfer between languages depend on them. Another advantage is the improvement of general knowledge and the reinforcement of reading for the remaining children [39].
«Open CV» (Open Computer Vision). It is a free library, targeting the Artificial Intelligence world. It was developed in 1999 and is currently the well-known library in artificial vision. There are established models on machine learning, although it is possible to optmize the library with deep learning [40]. It has different functionalities such as the movement detection and pose estimation. However, one of the most useful in education is detecting gestures that could be easily supossed as a significant advance for the autism spectrum disorder collective (ASD). One of the main features is the non-verbal communication either for expressing themselves or for gesturing. Thanks to major advances such as the OpenCV one, it would be feasible to work for improving the development and life equality in its first stages from the educational system [41].
«Tesseract OCR» is an optical recognition motor mainly used for in texts recognition and understanding of these writings through the machine. The fourth edition, now in beta version, is based on long short-term memory (LSTM). It allows the use of more than 116 languages ans scripts for 37. It was developed in the 80’s-90’s decade and financed by Google in 2006, making it in open-source. That was the time when it gained more acknowledgement and it was considered one of the OCR motors based on open source more accurate. When a text is recognised by Tesseract OCR, a computer can reproduce it and improve pronunciations and accents in different languages [42]. Tesseract OCR is supplementary to the use of OpenCV in respect of converting documents into images. By this way, it will be possible woking in a more simple mode. A code example to work with Tesseract OCR of Google would be:
//Import the libraries to use
import cv2
import pytesseract
//Use OpenCV to upload an image
img = cv2.imread('imagen.jpg')
//Grey scale for the image by using OpenCV
def get_grayscale(image):
return cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
//Remove the rumble of the image
def remove_noise(image):
return cv2.medianBlur(image,5)
def canny(image):
return cv2.Canny(image, 100, 200)
//Use Tesseract on the image
text = pytesseract.image_to_string(image, lang=es)
//Show the text extracted with Tesseract
print(text)
In this case, we have a text from the original document that it can be also used to other options.
To convert text into voice, we can apply the API from Google Translate by writing this lines of code:
//Import the Google API
from gtts import gTTS
//Import the os library
import os
//Indicate the texto to convert into voice and add the language
tts = gTTS(text=text, lang = 'es')
//Save the document
tts.save("texto.mp3")
//Open the file from the os library
os.system("texto.mp3")
Preschool (to the age of 6) is not compulsory. It is divided into two cycles of three academic courses:
The first cycle, from 0 to 3 years old, is taught in infant schooling and it is not offered by the Ministry of Education, Culture and Sport. The access to this cycle is not free.
The second cycle, from 3 to 6 years old, is taught in preschool centre and primary school. It is free accessible and offered by the Ministry of Educaation, Culture and Sport [43].
According to the Royal Degree 1630/06, the main goals of childhood education are: getting a comprehensive and harmonious development at the different levels (physical, motoric, emotional, affective, social and cognitive), in addition to project the learnings that make such development possible [44].
Schooling is an important educational benchmark, particularly in childhood education because apart from a requirement for professional training, it suposes the maturity of a set of values such as: independency, self-sufficiency, and the development of their own performance criteria [45].
The most of European countries have a high attendance rate for schooling in childhood education. See the following chart:
In Spain, the education rate from the age of 0 to 3 has increased in more than 10 points in the last years. In fact, Spain is now above average of the OCDE countries (25,5%). Moreover, in the second cycle of childhood, schooling is almost complete in Spain with percentages above 96% in all ages [46], as can be observed in the graph:
Notwithstanding the good results, there are countries such as Greece, Croatia, Luxembourg, Bulgaria, Austria, Finland and Netherlands, which are left under the 80% of schooling of 3 years old children. For instance, in Greece, in 2018, the gross schooling rate in childhood education did not exceed 50%. In comparison with the obtained results of the previous chart, the growth in one year is noticeable, although the increase has not proved to be too significant [47]. The results are a strong indication that schooling in childhood education is a ongoing problem, and therefore the education of very young children is negatively affected.
In accordance with urban teachers from official schools, the main difficulties in the process of schooling in childhood education are encountered in family environment. In fact, they pronounce some expressions such as: “…They have no interest for learning…, They are not interested in working…, They do not know anything…” [48].
Unfortunately, these social problems are not the only ones, but there are another factors involved in the process: single parents households, infant food allergies and material situation. To avoid problems of this type, the government of Poland has implemented an education management system, «AlgorithmWatch», which was developed by Asseco Data Systems. It is a system that works with algorithms based on AI and its function is assessing the elements to match children to the school where their needs can be covered. In the same way, this system is capable to provide an effective academic follow-up, facilitate communication between parents and an analysis of the assistance. The efficiency and profitability of this projects show good results that encourage schooling in childhood education [49].
Primary education is comprised of 6 compulsory and free courses. This educational stage is divided into 3 cycles:
Initial cycle: between the ages of 6 and 8 years.
Medium cycle: between the ages of 8 and 10 years.
Upper cycle: between the ages of 10 and 12 years.
The main goal is the learning of oral and written expressions and comprehension, reading and calculation. It is a synthesized meaning because primary education includes more concepts as direct form as transversal [50].
The goal of minimum teaching is ensuring a common training to all students in the Spanish education system and guaranteeing the validity of the corresponding titles, in accordance with Article 6.2 of the Organic Law 2/2006, on 3rd May, about education [51].
Learning difficulties are present in primary education. They are referred to a heterogeneous set of problems related to central nervous system. These are usually expressed in the basic disciplines of writting, reading and maths and learning in many curriculum areas as the second language acquisition [52].
According to the last PISA inform, Spain has a general punctuation of 481, that is, it is still below the average for the OECD, 487. For this reason, it is interesting to inquire what kind of problems have the children from primary education in maths and how Artificial Intelligence can do to settle their difficulties.
In the next chart, it is significant the evolution of the punctuations in maths, in Spain, in the last years of primary education stage.
As can be appreciated, maths is an unfinished subject in Spain. During the period from 2011 to 2015, this tendency has experienced a change. There is not a revealing data that claims that the evolution is positive or at least remained constant [53].
Artificial Intelligence is based on mathematical knowledgements and algebraic methods. With the use of these basic calculations can acquire most complete knowledges in the future. For example, learning about communications technology by developing a critical spirit at the received messages. Without information, there is no data and without data, there is no Artificial Intelligence. With a view to apply AI of any application, it is important to know the structure of data which are used to train the model and find the data if they are not available. On this stage, a whole world of possibilites is opened because children become conscious and this process acquires a more formal and theoretical character that the previous one.
«Smartick» is an application based on AI for learning mathematics and reading between the ages between 4 and 14 years, which includes the stage of primary education [54]. Smartick is the directed to skills development, comprehension, and process development. It also includes educational games that allow to improve cognitive abilities; memory, attention and reasoning. It is an application which can be used within and outside schools [55].
Smartick senses when students have cognitive problems. In mathmatics, this difficulty is called `dyscalculia´ [56]. In the next table, we analyse what cognitive areas are affected by dyscalculia and the proposed solutions by Smartick [57].
Affected cognitive areas | Cognitive training with Smartick |
Attention: concentration and attention to several stimuli at the same time | Games that require much attention: Chameleon, Camp Fire, Lost Objects and Double Assault |
Memory: solving mathematical problems, and retention of information | Nback games (working memory and fluid intelligence): Crazy Planet, Simon, Hologram and Memory |
Planning: sequencing important information to solve problems. | Applicatives: Chess for One, Four in a Row, Rush Hour and Bus Stop |
Processing speed: speed of stimuli to solve problems | Applicatives: Kiwi Run |
Likewise, Khan Academy is an online learning tool oriented to the teaching and support of the subject of mathematics [58]. It uses machine learning for determining student mastery [59]. It is also applied to secondary education.
The current research in this kind of applications is limited in respect of documentation of the usage of resources that students make because in the most of cases, these applications are used as outside school activities. Nevertheless, there are collected data of surveys to students beween the ages 5 and 9 years about Khan Academy in which the results reveal the perceptions, both professors and students, are improved because of the good academic results and they stimulate to learn maths [60].
As we have seen in the previous educational stages, there are many techniques based on Artificial Intelligence implemented in education. However, there is a limited number of proposals focused on learning of Artificial Intelligence by students and professors. It remains inexplicable why learning Artificial Intelligence in such important educational stages as secondary and bachelor is understimated.
Secondary Education aims to prepare students with the age between 12 and 16 years, whether to continue their vocational training and university studies or get started in the world of work. Among the new rules are highlighted: LOMCE 2013 and LOMLOE 2020 [61].
The Royal Degree 1105/2014, of 26 December, says that secondary education tries to achieve that students acquire the basic elements of culture, especially in its humanistic, artistic, scientific and technological aspects; develop and consolidate studying and working habits in students; preparing them for higher studies; entering them into the labour market; and training them for their rights and must in their lifes as a citicens [62].
Academic failure is an important educational indicative to bear in mind because education and academic preparation of students is the economic and social basis of a country. The number of graduates in secondary education in European Union is set in 85%, meanwhile Spain only manages 60% [63]. To verify it, in the next chart is represented, valuing in percentage terms, the academic failure in Spain in the last twenty years in comparison with Estonia, Finland and Netherlands:
If one of the main problems in compulsory secondary education in Spain is academic failure, it is important to review the causes and above all, if there exist measures based on AI that solve it. Some of the major causes of academic failure result from management for learning; task planning, how to solve difficulties, and teacher preparation [64].
Taking into account these aims and according to one of the issues present, we consider implementing computational thought, that is, the thinking process involved in the formulation of problems and solutions. By this way, the solutions will be represented to be processed by information processor (Cuny, Snyder y Wing, 2010) [65]. Alberto Valero (BQ Educación) says that computational thought is a key ability to digital development in classrooms. With that in mind, there is also the possibility of dividing problems, tasks planning, achieving results and testing if they are valid. Programming and robotics are suitable means to improve this type of thought [66].
With regard to teacher education, the University of Hong Kong, China, develops an initiative: Jockey Club AI for the Future Project (AI4Future). It is a project in collaboration with 14 professors, 17 directors and 6 secondary professors to create the first study plan of AI in secondary level. It also promotes the learning of AI between the students. 335 students and 8 secondary professors are part of the investigation. On the one hand, quantitative data reveal that the students have improved their competence, and indeed they have developed a more positive attitude to learn AI. On the other hand, this project encourages the autonomy of the professors for the effective conduct of this subject in the classrooms [67].
The University Carnegie Mellon of Pittsburg in Pennsylvania, has developed a project called Eduband that applies Artificial Intelligence in secondary education for professors to apply work methodologies based on student’s reactions and not only on the obtained results. The aim is to boost those subjects in which a student has more skills or he feels more comfortable, and not on those subjeccts in which he only gets good academic results [68].
Other interesting educational tool than can be used as a support in the process of learning is an application developed by Google «Socratic». It is focused on secundary education although it also encompasses to bachelor degree and university. The main objective of «Socratic» is offering support with homeworks.
As we can observe, there are some advances in the field of education that provide support to student population. However, it is also interesting question that we need to ask ourselves: how the usage of these applications can be controlled, while at the same time ensuring that children do not abuse it?
Bachelor degree is a postcompulsory secondary education, and therefore, is voluntary. The age period is between 16-18 years [69]. The three modalities of Bachelor degree in 2008, they have been increased in five [70]:
Science and technology
Humanity and social sciences
Arts in branches of music and performing arts
Arts in branches of plastic arts, image and design
General Bachelor degree.
According to the Organic Law 3/2020, 29 of December, bachelor degree is an educational stage that aims to fulfil the following objectives:
To reach intellectual and humanity maturity, knowledges, abilities and attitudes that allow students developing social functions and acquire certain competences.
To achieve indispensable skills to formative and professional future in the face of High education [71].
In this case, there is a close relationship bewteen vocational maturity and motivation oriented to academic learning. For this reason, it is needed to encourage to educational centres to find guidance strategies that advise students in order to boost their motivation [72].
Artificial Intelligence contributes to student motivation, developing the required competencies in a specific field of study. For example, the information technology department of the University of Southern California developed a programme of Bachelor degree in order to computer science based on AI. It is a project that uses gamification as an impulse. This proposal comprises two projections: «Computer games in the classroom», which uses the technology of video games to motivate students and «Pinball Project», which develops the software and hardware required so that students learn concepts about programming. The results of this experiment shows that, as well as increasing student motivation, it helps to train more computer scientists and, above all, best qualified [73].
In this educational stage, it is important that students have an idea about their professional future. In Spain, this topic is still an unresolved matter. According to a study made in 2018, 78 percent of the Spanish students in bachelor degree they do not know their professional future after studying bachelor [74]. A problem mainly caused by vocational maturity.
The application of Artificial Intelligence in vocational maturity has a positive impact in bachleron degree and higher education, as it shows a recent study carried out. This project allows, both professionals and student, to improve the knowledgement and verbalization of the existing skills, to compare competencies with objectives and needs in working life and to permit the access to sources of information updated about career services in order to generate future employment opportunities [75].
Artificial Intelligence also involves a greatest impact on higher education. Among the more used applicatives in higher education, we outlines chatbots, which are founded on natural language programming and destined to problems resolution about educational software, knowledge management in different contexts and streamlining processes [76].
Vocational Training includes learning oriented to insertion, reinsertion and job orientation with the aim of adjust the skills of future workers. In Spain, Vocation Training is divided into 3 levels [77]:
Basic Vocational Training
Training course intermediate level
Training course higher level
Dual Vocational Training is a modality in Vocational Training. In Dual Vocational Training, there are alternating periods between educational centre and company [78].
According to the Royal Degree 1147/2011, 29 of July, Vocational Training is an education stage that aims acquiring professional, personal, and social competences by students. Some of the objectives of VT are [79]:
To pursue the professional activity upon the general competence of the training programme.
To manage the professional career of the students, analysing training plans.
To use communications technologies applicable to occupational activity.
Apart from these objectives, the main purpose is motivating students, dealing with the problems of school absenteeism. For instance, getting a large number of college scholarships through agreements between educational centres and companies, increasing the responsability of business environment with academic education [80].
Artificial Intelligence is a challenge for Basic and Dual Vocational Training. The use of low-tech systems and “soft” skills for professional development, they can be digitized and intelligent, increasing the quality of requirements in staff training [81].
Currently, some initiatives are being developed which are directly connected to AI in education and technical and vocational formation. For instance, the education section of UNESCO in partnership with Ericsson launched the project Artificial Intelligence for Youth. It is focused on broadening the development of AI skills for young people. Its main objective is strengthening the capacities of trainers in order to empower young people in the development of applications based on AI. Because of the rapid evolution of technologies based on AI, it should be an alignment of the educational content in the VT programme and the needs of the labour market. This does become in a positive aspect for students because it is possible to teach the content which will help them to developm the skills demanded by enterprises [82].
The construction sector has recently experienced a decrease in the number of employees as well as the ageing of the best qualified workers. Therefore, carpentry skills did not change significantly. To improve this situation, it has been developed a teaching method based on machine learning with the aim of educating students in basic tasks of construction profession: sharpen, chisel and plane. In this case, AI is in charge of evaluating and measuring the movements of students, according to the previous patterns of qualified workers. Beyond the speed to correct the movements of students, the results show that this experiment ensures high precision in the professional activity [83].
«Knack» is an application of Silicon Valley based on proprietary gaming technology than combines a deep research and knowledgments about gaming theory, cognitive science, and the behaviour of artificial intelligence. The main objective of Knack is achieving the student self-awareness through games [84]. Knack helps candidates in Vocational Training programmes with more than 90 job opportunities in 17 different industry sectors [85]. Furthermore, Knack studies the user reaction under pressure, a way to prioritize tasks, learning from mistakes. Definitely, it is triumphant example that shows the acquired professional abilities and the demanded vacancies in the world of work, boosting Dual Vocational Training [86]. The main idea is to develop an initiative as Knack:
Only a 26% of adults wich work with university experience do agree in the theory that their education is important for their career and daily life.
The labour market demand and economic needs are changing rapidly.
The focus of educational systems in the university or professional preparation rather than fusing both [87].
More than 70% of studens will change their career [88]. It is a threatening situation because of economic consequences. It can also provoke a disorder in the familiar environment of the student, changing the mental state of the aforesaid student [89].
The revision of the study shows that the implementation of AI in education makes a positive contribution in relation to orientate the student to choose his professional career. It is a way to adaptate the capacities and circumnstances of the student to academic formation. Consequently, we can said that AI aids to job search [84]. To these results, the valuation of some professionals dedicated to higher education is added:
Knack’s gamification is an immersive and non-conventional smart recruitment experience. We use Knack’s innovative and engaging experience to challenge our business graduates in uncovering their problem solving and innovation skills and showcase their talents to employers, all while having fun1 (Erika Ortu, Gerente de capacitación, MIP Politécnico de Milano, Escuela de posgrado de negocios) [89]
Furthermore, the implementation of AI in Vocational Training is a form of promoting learning and formation in work centres. It is a way to encourage Dual Vocational Training practices by using different formation environments which are simulated and technologically improved. For instance, the intelligent workshop in Vocational Training schools in Germany [90].
University education is a learning process of higher education that takes place in an institution for searching, acquisition and construction of scientific knowledge [91].
Universities aim the formation of professionists of high scientific, humanistic, and technological quality that contributes to the development and wellness of a country [92].
Most problems involve cognitive difficulties; specific training and learning problems, attention deficit hyperactivity disorder (ADHD) and autistic disorder [93]. These are health problems that are a challenge for educational institutions in childhood, pimary, secondary, and special education. However, with techniques based on Artificial Intelligence, as we can see in the paragraph of special education, this picture could be changed, rendering these people, who really deserve it, access to a well-rounded education.
Other difficulty in university education is acquiring the professional skills to find a job, and put these acquired abilities into real practice, preferably in the country of origin of the student. To do it, the University of Helsinki, Finland, develops a project with the metropolitan area of Helsinki to map the future curriculim of students by using Artificial Intelligence to analyse the needed competences in each area at the country [94]. From now on, we should reflect about this question: Is there exist any worry for the future of students or the economic growth of the country? It would be useful to highlight that the use of Artificial Intelligence has to be associated with human data protection.
Evaluation is a controversial issue in learning process because in last years it has been considered as a process capable to recognize new educational approaches, different uses, methodologies of implementation and objectives. The main aim is to get a more equitable mark for students [95].
CBT systems arealternative methods to traditional methods of evaluation. These systems involve the use of communications technology to do exams [96].
This study is focused on the analysis of three of the most known evaluation systems based on CBT:
Linguaskill. It is an online flexible tool that evaluates English skills for business, industry and commerce. It serves as a support for workers to feel them confident to communicate in international business environments. Linguaskills is the new version of Bullats, created by Artificial Intelligence technology [97]. With Linguaskill, it is possible to create certificates and tests in the University of Cambridge through the department of Cambridge Assessment English-ESOL [98].
Prometric. It is a leading supplier of assessment solutions based on technology for certification organizations, academic institutions, and the most known government agencies in the world [99]. With Prometric, it is possible to do IREB exams to acquire a deeper understanding of business analysis and requirements engineering [100].
Pearson Vue bets on exams that aiming to empower to professionals to certificate individuals which advance in their communities at global level. Pearson Vue develops technologies that lead through essential credentials in a virtual way in each industry [101]. Thanks to Pearson Vue it is possible to have access to certifications such as Cisco Networking Academy, which offers specialisations in industry, to design and network support [102], as well as Juniper, Adobe and Oracle.
CBT systems have some advantages [103]:
Availability. These kind of tests can be done throughout the year. It is not necessary that students travel to a specific location to do an exam.
Adaptability. The level of difficulty of the test is established according to the his knowledgment, capacity and answers.
Integrity. Online monitoring and security functions make CBT systems in secure way of evaluation.
Scalability. The exam can be done simultaneously by several candidates.
Automatic qualification. Minimizing human errors and the inconvenience of assigning evaluation tasks for professors.
Inclusion. Integrating characteristics such as braille keyboards, strong tools, screen readers, voiceto-texts systems, and text applications.
However, these systems also present disadvantages:
Security problems with Pearson Vue. In the credential management system, it was detected a Malware. This system is mainly used by students to have access to professional certifications. A malicious user had access to private information from other users, which is required to acquire a certification; credit card number, and personal identification number [104].
Invasion of students’ privacy. One of the dangers for doing the tests is identity theft. To avoid it, some centres put into practice facial recognition, although according to Spanish Data Protection Agency, it needs forced guarantees [105].
Fraud. Certifications have become essential to career advancement. Because of it, frauds and scams have been increased in last years. Despite of CompTia, other organization which valuates professional skills for technological information industries [106], it estimates that the number of frauds in these kind of exams is about 5%, some experts in the area say that this is an increased problem and we should be in permanent alert [107].
Even though remote systems taking precedence in the academic values of students, experts in education also highlight the importance of traditional paper assessment [108].
Providing teachers the opportunity of having a communicative relationship with students, getting the best learning results.
Written communication offers transparency for students about their level of understanding.
Comprehension of a concept rather than answering the correct response.
Apart from these advantages, traditional evaluation systems have also disadvantages [109]:
Time. It requires more time and high costs in materiales to evaluate big groups of students.
Results. Delay in the processing and interpretation of results.
Limitations. In written assessment, testings are limited and they offer the same kind of resource.
To recover this data graphically, we can see the next tables:
Traditional evaluation systems
Pros | Cons |
Communication between teachers and students | High costs in materials and time |
Providing transparency in comprehension level | Limited tests and lack of resources |
Deep understanding of concepts | Delay in the interpretation of results |
CBT systems
Pros | Cons |
Possibility of doing tests throughout the year | Access to private data of users without prior agreement |
Setting of level difficulty | Identity theft |
Online monitoring | Fraud in certifications |
Simultaneity | |
Automatic qualification | |
The inclusion of material |
Chatbots are the most useful Artificial Intelligence tool to apply in education. According to Sánchez and Ayala (2018), chatbots are computer programmes with which users can maintain conversations, either by text or voice [110]. Although they are not used as much, we can find them in differents working areas, chatbots still not have a big impact on education. For instance, they can be used to offer a more individualized attention and also for the resolution of doubts about the syllabus when children are out of the scholar schedule. It can help as the academic development of student as proffesional development of teachers, who invest less time to resolve doubts [111].
If we think about AI in other countries, we highlight two territories that appeared in the first position of PISA; Finland and Sweden. These Nordic countries have two opposite positions in respect of their curricula, that is, to the whole of knowledges that a student should have to get an academic title [112]. Finland gives the possibility of developing AI strategies to schools, despite of not notify in its curricula. Meanwhile, in Sweden, the problem itself is not the infrastructure but rather the competences. They are present differences between schools and even classes, because in some cases the use of AI is omitted, although it is clear that AI is a tool to develop by all students.
In the last years, some technologies have been developed, for instance: «Squirrel AI», based on AI. It is a Chinese technological enterprise which offers the option of teaching adapted to students throughout online courses and scholar assitance. It also provides after-school classes to reinforce learning and help them to pass exams. Currently, in China, it is included in almost 2000 educational centres and it services to more than 2 million of students [111].
It is important to highlight that in an interview with MIT Technology Review, the CEO of Squirrel AI says:
“Cuando la educación mediante IA prevalezca, los profesores humanos serán como un piloto"2 [68]
While it is true that AI provides facilities that improve the process of teaching, its development is costly, so it is not accessible for everybody [113]. In a way, the implementation of AI in education may be a widening social gap, especially if sources based on AI are used in extra-curricular time and they are paid by families. However, the scenery changes if AI was implemented in the curricular of educational systems of the countries and the government will was the responsible of subsidizing all the material for all students, without exceptions. A clear example of that is South Corea, where the government has decided to provide didactic material in schools and kindergartens the Alpha Mini robot, that teach them how to dance, sing and even tell jokes. This initiative was developed with the idea of knowing how to manage Artificial Intelligence and tools related with it, because AI will be important for the future [114].
We thank to organizations and personalities that make possible this project by collaborating on updated information and testimonies which enhance the investigation value.