A Model of Personalized Context Aware E-learning Based on Psychological Experience

Abstract- The use of context aware in the model of personalized e-learning has brought a new passion for users as an alternative to learning. Personalized context aware e-learning can provide adaptive learning patterns that can be tailored to the needs, the circumstances and the behavior of users. Along with the continued development pervasive and ubiquitous computing, the personalized context aware e-learning continues to be developed. However, models of the personalized context aware e-learning developed still focus on a wide variety of contexts, such as physical contexts, explicit contexts, and context related to the physical environment of learning. Additionally, the development of the models of the personalized context aware e-learning by considering psychological experiences as a context is still limited. In fact, the engagement and interest in learning are determined by the psychological condition of users. This research will develop a model for psychological experience of the personalized context aware e-learning. The psychological experience is based on the theory of flow consisting of anxiety, boredom, and optimal condition measured naturally when users are interacting with e-learning. Furthermore, it becomes one of the strengths of this research. Psychological experiences are measured using a machine learning method using data obtained from users’ behavior saved in the server log.

Index Terms— Awareness, context, context aware e-learning; e-learning; flow experience; personalized e-learning.

I.     Introduction

Learning that is based on Information and Communication Technology known as e-learning is currently being developed in order to provide learning alternatives, equity and access to education. Nevertheless, there are still issues that are often encountered in the implementation of e-learning system at this time. Those issues include: pedagogical aspects, lack of user collaboration, limited context of social campus, technological gap, the emotional state of the user, limited digital content according to user interest, and the loss of learning context [1].

The increasing needs of the users that are diverse encourage an e-learning system to be designed and built following the natural condition of users. The virtual learning environment is developed as much as possible to meet the conditions and situations of the users. Therefore, the concept of e-learning is concerned with learning context, giving more attention to the needs of users, as well as adopting previous experiences [2].

Points presented previously involving a variety of issues related to the situation, the learning context, and changes in teaching methods have encouraged the development of e-learning models directed and aimed at following personal needs of the users in the form of personalized e-learning. Parameters of the context are in the form of needs, characteristics, situations, and conditions of the users to gain awareness on learning and represent a context aware e-learning [3]. Unlike conventional e-learning, a model of personalized context aware e-learning provides adaptive solutions based on the conditions that can be done in various contexts [4].

Along with the continued development of pervasive and ubiquitous computing, the personalized context aware e-learning continues to be developed, including in the form of mobile learning [5]. However, the existing models of personalized context aware e-learning developed still focus on the presentation of learning materials with a wide variety of contexts, various methods, and learning environments. Development of the model of the personalized context aware e-learning by considering the psychological aspects of the learners as a reference to the context and specific learning model is still limited [6]. In fact, the engagement and interest of learners in the learning process is determined by psychological conditions of learners [7].

The most common psychological experience in learning is a feeling of anxiety, boredom, and optimal conditional experienced by learners. The third condition is based on a theory which states the conditions of emotional flow, cognitive and motivation of learners [8], [9]. It consists of optimal conditions (flow state), anxiety and boredom. Thus, this study proposes a model of the personalized context aware e-learning based on the psychological condition which focuses on the balance between presentation levels of challenge/problem and skill/learning materials. Psychological experience will be the context in generating awareness regarding adaptive response in the form of e-learning to the learners.

II.   Literature Review

A.      Related Works on Context Aware E-learning

Many studies related to the model of the personalized context aware e-learning have been done in order to improve e-learning model at this time. Besides using the terminology of the context aware e-learning, several studies have also used the terms of recommender system, as well as personalized and adaptive context aware e-learning. Generally, these studies focus on how to provide personalized learning using a variety of contexts. The rapid development of pervasive and ubiquitous computing, the models of the context aware e-learning generally involve contexts related to the physical environment. The context of location, time, noise, lighting, temperature, and physical environments are more widely used to provide recommendations tailored to available physical sensors.

Research in the context aware e-learning using RFID as a context sensor location has ever been undertaken in order to provide recommendations in the form of position detection of physical learning resources and closest learning peers. Physical learning materials are used to support materials in the exploration of knowledge, while the detection of the closest pers is intended to provide information about peer learning [10]. The same action is also done using GPS sensors to seek the location of learning peer and learning resources nearby [11]. Shu-Lin Wang uses the context location with the RFID and personal learning profile for recommending appropriate learning objects [12]. In the context of other locations, Scott et al, add infrared sensors (IR), RFID and GPS on the model proposed by the recommendation materials, peer and learning tasks [13].

The context of times, locations, and activities are often used to provide recommendations for personalized services to users. Time is marked with a timestamp to a location or activity. Locations are associated with outdoor, indoor, public places, as well as private places. While activities may consist of rest, waking, working, walking and so on [14]. Zhao et al, use the context of time, location, quality of the network and the type of device to provide adaptive learning content through a mobile learning [15]. In their research, Isabela et.al., use the context of time and location combined with the cultural background of a system to present recommendations and language learning material used [16].

Social context through feed back from social media tag has been used to generate recommendations for appropriate learning materials, peer learning, and tutors that suit users [17]. Based on this context, e-learning models are built and integrated with several social media applications to gain user input when interacting with social media via the feedback or tag of materials, and tutors.

User profile was ever used to connect learning material to profile of learners using matching context [18]. Luo., et.al conducted research in a model of personalized e-learning with recommended learning materials for students using the context of individual preferences [19]. In their study, the model provides learning materials relevant to learners from a set of available material. Nguyen and Phung use the context of this profile to present learning materials and pedagogical approaches appropriate for learners [20]. Meanwhile, the models and the characteristics of learners are also proposed to classify learners who have the same preferences to learning materials. Learners who have the same profile model get the same learning materials. Grouping is done by using the context Bayes learning styles [21].

Other studies generally propose a model of personalizd context aware e-learning by creating a context grouping used in personalized e-learning. Shudana classifies contexts to adapt learning materials into three classifications, namely: situation context, domain context and context activity [22]. Situation context is related to hardware, software, network and location of learning. Domain context is related to topic and language used. Additionally, activity context is related to interests and goals of the learners. Gallego., et.al offer a model to acquire learning materials based on the social context (associated with the interest of students in courses offered), the location context (associated with the time and the geographical location of learners, country and language), as well as user context (related equipment / devices used by the user in accessing the learning materials) [23].

Grouping of the context aware e-learning is also done by Capuano., et.al to describe the model context aware e-learning using an approach of ontology. In their study, five classifications are proposed, namely educational context (associated with the state and level of education), course subject context (associated with the subject being undertaken), methodological context (related learning approaches: self learning, synchronous, asynchronous, blended learning, formal learning), instructional context (associated with active learning, collaborative learning, game-based learning, inquiry learning), as well as the technological context (associated with the constraint of the device and the constraints of the network) [24].

Das divides the contexts of the parameters into four classifications, namely: profile context (related to the learner information profile, level of expertise, personality type), preference context (associated with learner intention, learner approach / preference, learner style), infrastructure context (associated with learner situation, network, and device) and learning context (related to the learning state and a more comprehensive level of learners) [3]. The same work is also done by Verbert., et.al who describe the results of exploration and survey of e-learning context aware. Research suggests a wide variety of contexts used in e-learning, such as computing context (network, hardware, software), location, time, physical condition, activity, resource, user, and social relations [25].

Other studies related to the grouping of the contexts in the model of context aware e-learning are focused on the users and the user interaction with e-learning. Users are represented by internal and external profile, while the interaction with the e-learning is represented by the user interaction with the application/tutor and other users. While the recommendation is associated with learning object profile, service profile and tutor profile [26].

B.      Related Works on Flow in Learning

The implementation and application of the flow theory and technology acceptance model (TAM) were developed by Liu and Yuan in the online e-learning users [27]. This study aimed at looking at the user’s acceptance behavior that was integrated into the web-based streaming e-learning. This study used a text-audio learning materials, audio-video, and text-audio-video. Variable flow experience used is concentration, while TAM uses perceived parameters of usefulness, perceived ease of use, and attitude which have impact on the intensity of learning. The results of this study reported that the emphasis of the presentation of learning materials could build up a concentration of learners, so that the needs of learners should be carefully considered when designing e-learning.

Another study has tried to apply flow experience into a face to face learning environment. This study focuses on the relationship of flow experience with multimodal learning environment. The purpose of this study is to find out more about the flow experience of learners in the problem based learning by involving multimodal learning environment. Another aim is to increase multimodal learning environment to understand the environment that can enhance the flow experience of learners. Multi-modal learning environment in this study was conducted in the classroom by involving a wide range of sources of learning materials.

P.I. Santosa proposes a conceptual model of flow theory in utilizing e-learning [28]. This study uses two variables subjective challenge and skill. Flow experience is seen as correspondences between challenge and skill. In this study, the challenge is represented as lectures and web navigation, while skill is represented by the web prior knowledge and experience. Furthermore, the symptoms are represented by the perceived flow of control and concentration. In this study, symptoms of flow affects are proposed to influence the performance of learners. This study reveals that the design of e-learning should motivate students to stay focused on the materials presented so that there is a challenge for how to design e-learning to accommodate the characteristics of different students.

A similar work has ever been applied in viewing student engagement in the online tutorial. In this study, flow is represented by the size of the web page (challenge) and prior knowledge (skills). Both of these variables are positively correlated with the ease of navigation related to attitude that is related to student engagement. This study also involves two other variables such as the perceived benefit that positive things are gained by students during using e-learning, while the perceived costs are the things that give negative impacts on the learner when interacting with e- learning [29]. The study concluded that regarding technology use in online learning, the student engagement is influenced by various factors related to the design of online learning

III. Theoretical Background

A.      Context Aware Computing

The term of context-aware was first introduced by Shilit and Theimer in 1994. According to their context is the location of use, a collection of people, close objects, and the change of the objects over time [30]. Context is defined as any information that can be used in order to characterize something on an entity which may be a person, a place or a physical object that is computing [31]. According to Dey, Context is any information that could be used in order to characterize the situation in an entity. The entity can be either a person or object that is relevant in user interaction with the application itself [32].

Context awareness is the use of contexts to provide information and task-relevant interactive services between users and the elements in the surrounding environment. A system is context aware if the system uses contexts to provide information and services relevant to users, where relevancy depends on the user’s activities [33]

B.      Context Aware E-learning

In the previous section, it has been described about the context aware based applications. In the case of e-learning, context is defined as the current situation with regard to the learning activity [34]. A context in e-learning may be prior knowledge, learning style, learning speed, current activities, learning objectives, time availability, location, and other interest [35]. A context in an e-learning system is used for personalization, recommendations and adaptation of learning materials in accordance with the interests, circumstances, environment, and learning style of the user [22], [36]. Therefore, in e-learning environment, a context can be seen from a user perspective by considering various factors that influence learning style. In the context-based adaptive learning system, it has basically three important stages that must be performed, namely acquisition context, context modeling and adaptation context.

C.     Personalized E-learning

Personalization concept is the process of changing or adding something to the object so that it matches the needs of an individual. In the context of the e-learning environment, personalization has become a very important topic because the learning process is no longer done the same for all individuals who have ways, preferences, and interests. Therefore, it becomes a very important thing if an e-learning system can provide materials, paths and learning approaches according to each student’s needs and expectation [37].

Personalization of e-learning is defined as a learning approach that facilitates and supports individual learning where each user has a path and learning services according to his/her needs [38]. In general, the process of adaptation in the personalization of e-learning can be done in three processes, namely: Selection, Sequencing, and Presentation [39]. Selection regard to the selection of learning materials in accordance with the levels. Sequencing is concerned with how the learning material or learning object which is produced from a selection is prepared in accordance with the user’s individual learning paths and pedagogical approaches were undertaken. Meanwhile, the presentation is concerned with how learning materials are presented in various forms of media, size, and others.

Adaptation to the personalization of e-learning is based on components of the personalization that include parameters, conditions, and contexts that describe the characteristics of learners. These characteristics are obtained through data classification of learners from a series of surveys that are entered to the e-learning system as well as from data obtained from the interaction of learners with e-learning system. Examples of contexts / parameters that can be used as components in the personalization of e-learning can be seen in the Table 2.

The contexts / parameters will determine about type of personalization. As much as possible to match with the given context/parameter. Based on Table 2, there are three types of contexts : personal, abstraction, and situation, These type of contexts are supported by several parameters that describing the contexts.

Table 2

Components of flow

D.     Flow Theory

Flow theory was first introduced by Mihaly Csikszentmihalyi who used the terms of flow to represent the optimal experience of someone to focus on his/her engagement in an activity [40]. Some of the elements and characteristics of flow conditions indicate improvement and motivation in learning. Hoffman and Novak summarize the elements of the flow from several previous studies into flow antecedents, flow experience, and consequences flow as shown in the Table 1.

Table 1

Components of flow

Although the flow is built by several complex variables, but the skill and challenge are the two most important variables [41], [42]. Therefore, the flow experience is often depicted with these two variables are involved. In general, the flow theory puts forward three conditions, namely the optimal conditions (flow state), anxiety and boredom. The optimal condition is reached if the skills are in line with the given challenge, nervous conditions occur if skills are low and challenges are high, while the condition of boredom occurs if skills are high while challenges are low.

IV.                The Proposed Model

A.      Model Development

Related studies as discussed earlier have given an overview of the context and recommendation (awareness) used in the model of context aware e-learning. However, the models that have been presented are still largely associated with the classification of contexts with the various recommendations presented. The available models do not fully display context and recommendations in a generic way, particularly on the context aware e-learning. It includes models that have not yet covered the process that should be present as process of the acquisition context and also adaptation process, especially in the process of evaluation and measurement of e-learning model.

The studies that have been done suggest that context is multi-dimensional. In the model of context aware e-learning, it can be involved one or a combination of some contexts. Similarly, with the recommendations presented, they may involve one or a combination of several recommendations. Everything is tailored to the needs, situation, goals and learning scenarios conducted. The representation contexts as presented in previous studies can be summarized in Figure 1.

Figure 1: Context representation

The contexts used in the studies that have been conducted generally remained static in accordance with the information entered. If there is any change in conditions of the context, it is caused more by learning environments and physical devices (external users). The contexts showing behavioral, emotional state, motivation, engagement and user interest while interacting with e-learning are still limited. On the other hand, information about psychological condition indicated by the user behavior in e-learning is a very important aspect.  Involving contexts in e-learning can support supervision of user’s e-learning from the affective point of view.  These contexts can not be obtained implicitly or explicitly, but through a process of inference through data patterns as a result of user interaction with e-learning. Based on this information, the context acquisition process can be classified in three ways, as shown in Figure 2.

Figure 2: Context acquisition

The representation of context on the model of the context aware e-learning determines the kinds of recommendations (awareness) generated. The recommendations presented in studies that have been previously conducted are mostly related to the selection of teaching and learning materials. However, recommended places and time of the study, the location of the source of learning, and learning peer have always been the focus, especially in mobile learning models. These recommendations are generated based on the adaptation of one or a combination of several contexts identified. As well as the context, the models that have been done do not represent awareness that is generic from a model of e-learning. In general, the recommendations (representation awareness) are shown in Figure 3.

Figure 3: Awareness representation

B.      Conceptual Model for Context Aware E-learning

The main components that must be completed in the model of the context aware e-learning are the presence of three main dimensions, namely user, context, and awareness. The user is primarily concerned with the learner involved directly in the model built. Context is based on state of the art e-learning context aware. In the previous section, it is classified into more generic into the internal learner context, external learner context, learner interaction, and learning context. While the dimensions of awareness generally include adaptation related to learning materials and related to the learning activity. Based on this framework, one user can associate with many contexts and many types of adaptation. Or, many users can associate with one context and one or many type of adaptation. The situation depended on the needs of the model. Furthermore, another important component in building the model of the context aware e-learning is the process of the acquisition of context and process of adaptation awareness. The conceptual model of the e-learning context aware in this study can be seen in Figure 4.

Figure 4: Conceptual model context aware e-learning

Users referred to in the conceptual model are learners who received the e-learning services. Users such as tutors, academic managers, and system administrators are categorized as part of their e-learning system with the main role as facilitators of services to the learners. Other dimensions on this model can be described as follows:

• Context

In a context aware application, in general, context is defined as all kinds of information that can be used to characterize / illustrate the situation of an entity [32]. An entity can be a person, place, time, or another object that is considered relevant to the interaction between users and applications. In the model of the personalization of the e-learning context aware, this context is divided into four as an internal learner context, external context learner, learning context and interaction context. Internal learner context is associated with the user context includes profiles, background, culture, language, preferences, activities, learning style, and other psychological conditions. External learner context is related to the physical environment of learning such as brightness, noise, the device being used, network, location, time and other physical environmental conditions. Learning context is related to the learning instruction, pedagogy, and course type. Additionally, the interaction context is primarily concerned with social relations as well as user interaction with other systems.

• Acquisition and Adaptation Engine

Acquisition / Adaptation engine for e-learning can be described as a unit process that requires input from the learner contextual information and makes a recommendation / awareness to the learner [43]. Input is obtained through the acquisition context as presented in Figure 2. From some of the literature review undertaken, the approaches can be identified into two: acquisition / adaptation engine that is implemented on a model of the context aware personalisai e-learning, namely: (i) acquisition / adaptation rules; (ii) acquisition / adaptation algorithm.

Through the approach of acquisition / adaptation rules, acquisition and representation context, recommendation and personalization are obtained through a state sentence structure (if / else / then statements). While the approach to acquisition / adaptation algorithm, acquisition and representation context, recommendation and personalization are done by applying various types of algorithms, such as heuristic algorithms, intelligent approaches, matching algorithm, artificial algorithm, similarity algorithms, decision-based algorithms, and methods related to machine learning.

• Awareness Representation/Type of Adaptation

Broadly speaking there are two categories for the forms of representation awareness (adaptations) of the model of the personalization of the context aware e-learning, which is associated with the adaptation of learning materials, and the second is related to the adaptation of the learning activity [44].

Adaptation is related to learning material consisting of three forms, namely: (i) selection; (ii) presentation; (Iii) navigation and sequencing. Adaptation by selection with regard to the selection of instructional materials / problem right based on selection criteria derived from the learner contextual information. Adaptation based presentation is how to show learning material based on the device being used, the location and time, as well as the format of the material available. While adaptation related to navigation and sequencing, recommend structuring the possibility of navigation and sequencing of different learning materials that are connected to one another to create personalized learning paths.

Adaptation related to the learning activity consists of four forms, namely: (i) general adaptations, this adaptation automatically provides adaptive learning activities based on the criteria of context; (ii) feed back and support (scaffolding), this adaptation is generally recommended time and appropriate learning activities; (iii) navigation to locations, this adaptation is primarily concerned with the event, and learning resources; (iv) communication and interaction, this adaptation provides recommendations with regard to collaborative learning activities such as peer information, tutors, etc.

• Evaluation the Model

Evaluation of the model of context aware e-learning can be conducted in parts and all parts of the model. Evaluation of part of the model especially related to accuracy value of context identification or accuracy during type of adaptation. The accuracy can be conducted using prediction results compared to questionnaire results. If the model implements the maching learning methods, the accuracy can be reached through confusion matrix. Meanwhile, the usefulness, effectiveness, and another usability can be done for the whole model after be implemented in the real situation. One of the evaluation and measurement tools for the whole system can use the user experience questionnaire.

C.     Contextual Model for Context Aware E-learning Based on Psychological Experience

A contextual model of the context aware e-learning in this study is based on the psychological experience of learning. It is based on state of the art of previous studies showing that the context-context related to user shows the variation and the dimensions are diverse. While the contexts related to the physical environment, location, time, software and hardware, as well as common human activities do not show a lot of variations. Therefore, the dynamics and variations user context is high enough. The addition of the psychological context in the model of the context aware e-learning is very important and supports the affective side of learning. On the other hand, the psychological experience shows one very important aspect in learning. The learning process with attention to psychological conditions can affect the learner’s interest and engagement in learning.

Psychological condition identified as context in this study is represented based on the theory of flow. Someone on the flow conditions is at a high concentration so that there is no room for other thoughts or distractions.

Figure 5: Contextual model contex aware e-learning

Although the flow is built by several complex variables, but the skill and challenge are the two most important variables [41], [42]. Therefore, the flow experience is often depicted with these two variables involved. Therefore, the balance of skill and challenge/problem is the focus on the process of adaptation / personalization models. In this contextual model, the skill is supported by the personalization of learning material based on learning style context, while personalization challenge/problem is based on the experience of the flow context. The contextual model can be seen in Figure 5.

Figure 6: Acquisition and adaptation process

The contextual model processes the acquisition of the context by inference. Learning style and flow experience are inferences based on learner behavior when interacting with e-learning. The context is acquired based on attributes / parameters of log data on the server. Furthermore, these attributes are extracted, represented into context using the acquisition engine. For each learner with the identified context will be given in the form of personalized learning and problem. This process is performed by the adaptation of engine on the model. As an illustration, the context acquisition and the type of adaptation processes can be seen in the Figure 6.

In term of psychological experience, type of adaptation of the contextual model, aims to monitor and control that the skill and challenge are in balanced condition (flow) as can be showed in Figure 7. Therefore, this contextual model provides selected learning material according the learner style and selected challenge according to the flow state.

Figure 7: Flow states

V.  Conclusion

The model of the context aware e-learning based on psychological experiences has been presented in this study. In general, this model is developed based on the conceptual and contextual model. The conceptual model provides the framework and processes that can be used in applying the model of the context aware e-learning. Meanwhile, the contextual model is based on the psychological experience based on the learning style and flow experience context. The combination of these two contexts really supports the creation of a balance of skill and challenge for achieving optimal learning. Therefore, this research is expected to contribute as the guideline for the development and improvement of the model of the context aware e-learning.

References

1. T. Kenan, C. Pislaru, and A. Elzawi, “Trends and Policy Issues for the e-Learning Implementation in Libyan Universities,” Int. J. Trade, Econ. Financ., vol. 5, no. 1, pp. 105–109, 2014.
2. L. Uden and E. Damiani, “The future of E-learning: E-learning ecosystem,” 2007 Inaug. IEEE Int. Conf. Digit. Ecosystem. Technol. (IEEE DEST 2007), pp. 113–117, 2007.
3. M. M. Das, “STATIC CONTEXT MODEL FOR CONTEXT AWARE E-LEARNING,” vol. 2, no. 6, pp. 2337–2346, 2010.
4. [4]              R. Sunitha and G. Aghila, “LEMON : THE LEARNING EFFICIENCY COMPUTATION MODEL FOR ASSESSING LEARNER CONTEXT IN E -LEARNING,” Int. J. Integr. Technol. Educ., vol. 2, no. June, pp. 23–32, 2013.
5. A. Manzoor, H.-L. Truong, and S. Dustdar, “Quality of Context: models and applications for context-aware systems in pervasive environments,” Knowl. Eng. Rev., vol. 29, no. 2, pp. 154–170, Mar. 2014.
6. G. Zhang and Z. Cheng, “A WWW-based learner’s learning motivation detecting system,” 2003.
7. C. Whitson and J. Consoli, “Flow Theory and Student Engagement,” J. Cross-Disciplinary Perspect. Educ., vol. 2, no. 1, pp. 40–49, 2009.
8. J. Nakamura and M. Csikszentmihalyi, “The concept of flow,” The handbook of positive psychology. pp. 89–105, 2002.
9. D. J. Shernoff and E. Rowe, “Measuring Flow in Educational Games and Gamified Learning Environments Increasing Engagement in Learning through Serious Educational Video Games Theoretical Foundation : Flow Experiences and Their Relationship to Learning,” pp. 2276–2281, 2012.
10. M. Derntl and K. a. Hummel, “Modeling Context-Aware e-Learning Scenarios,” Third IEEE Int. Conf. Pervasive Comput. Commun. Work., pp. 337–342, 2005.
11. S. Martin, R. Gil, E. S. Cristobal, G. Díaz, M. Castro, J. Peire, M. Milev, and N. Mileva, “Middleware for the development of context-aware applications inside m-learning: Connecting e-learning to the mobile world,” 4th Int. Multi-Conference Comput. Glob. Inf. Technol. ICCGI 2009, pp. 217–222, 2009.
12. S. Wang and C. Wu, “Expert Systems with Applications Application of context-aware and personalized recommendation to implement an adaptive ubiquitous learning system,” Expert Syst. Appl., vol. 38, no. 9, pp. 10831–10838, 2011.
13. K. Scott, S. Member, R. Benlamri, and S. Member, “Context-Aware Services for Smart Learning Spaces,” vol. 3, no. 3, pp. 214–227, 2010.
14. [14]              A. Moon, Y. Choi, and B. L. G, “Context-Aware User Model for Personalized Services,” pp. 858–863, 2008.
15. X. Zhao, F. Anma, T. Ninomiya, and T. Okamoto, “Personalized Adaptive Content System for Context-Aware Mobile Learning,” vol. 8, no. 8, pp. 153–161, 2008.
16. I. Gasparini, M. S. Pimenta, and J. P. M. De Oliveira, “Combining Ontologies and Scenarios for Context-Aware e- Learning Environments,” SIGDOC, ACM, pp. 229–236, 2010.
17. L. Hu, Z. Du, Q. Tong, and Y. Liu, “Context-Aware Recommendation of Learning Resources Using Rules Engine,” in IEEE 13th International Conference on Advanced Learning Technologies, 2013, pp. 181–183.
18. P. Moore and H. V. Pham, “Personalized intelligent context-aware e-learning on demand,” Proc. – 2012 6th Int. Conf. Complex, Intelligent, Softw. Intensive Syst. CISIS 2012, pp. 965–970, 2012.
19. J. Luo, F. Dong, J. Cao, and A. Song, “A context-aware personalized resource recommendation for pervasive learning,” Cluster Comput., vol. 13, no. 2, pp. 213–239, 2010.
20. L. Nguyen and P. Do, “Learner Model in Adaptive Learning,” World Acad. Sci. Eng. Technol., pp. 395–400, 2008.
21. D. Zakrzewska, “Building context-aware group recommendations in E-learning systems,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 6922 LNAI, no. PART 1, pp. 132–141, 2011.
22. K. M. Sudhana, “An Ontology-based Framework for Context-aware Adaptive E-learning System,” pp. 0–5, 2013.
23. D. Gallego, E. Barra, S. Aguirre, G. Huecas, and T. Escuela, “A Model for Generating Proactive Context-Aware Recommendations in e-Learning Systems,” 2012.
24. N. Capuano, M. Gaeta, S. Salerno, and G. R. Mangione, “An ontology-based approach for context-aware e-learning,” Proc. – 3rd IEEE Int. Conf. Intell. Netw. Collab. Syst. INCoS 2011, pp. 789–794, 2011.
25. K. Verbert, N. Manouselis, X. Ochoa, M. Wolpers, H. Drachsler, I. Bosnic, and E. Duval, “Context-aware recommender systems for learning: A survey and future challenges,” IEEE Trans. Learn. Technol., vol. 5, no. 4, pp. 318–335, 2012.
26. S. B. Lau and C. Lee, “Context Aware Reference Model : Architecture and Implications for Adaptation of Learning Activities,” in 4th Intl. Conference on Mobile Technology, Applications and Systems, 2007, pp. 623–627.
27. S. Liu and C. Yuan, “Applying The Technology Acceptance Model and Flow Theory to Online E-Learning Users’ Acceptance Behavior,” Int. Assoc. Comput. Inf. Syst., vol. VI, no. 2, pp. 175–181, 2005.
28. P. I. Santosa, “Model Konseptual Pemanfaatan Teori Flow Dalam eLearning,” J. Nas. Pendidik. Tek. Inform., vol. 1, no. 1996, pp. 24–28, 2012.
29. P. I. Santosa, “Student Engagement with Online Tutorial : A Perspective on Flow Theory,” Int. J. Emerg. Technol. Learn., vol. 10, no. 1, pp. 60–67, 2015.
30. B. Shilit and M. Theimer, “Disseminating active map information to mobile hosts,” IEEE Netw., vol. 8, pp. 22–32, 1994.
31. P. . Brown, J. . Bovey, and X. Chen, “Context-aware Applications : From the laboratory to the marketplac,” IEEE Pers. Commun., vol. 4, no. 5, pp. 58–64, 1997.
32. A. Dey and G. Abowd, “Towards a Better Understanding of Context and Context-Awareness,” 1999.
33. T. Mantoro, “Distributed Support for Intelligent Environments,” The Australian National University, Australia, 2006.
34. R. Luckin, Re-designing Learning Contexts: Technology-rich, Learner-centred Ecologies. 2010.
35. Z. Yu, X. Zhou, and L. Shu, “Towards a semantic infrastructure for context-aware e-learning,” Multimed. Tools Appl., vol. 47, no. 1, pp. 71–86, 2010.
36. R. Prabagaran, R. Guna, and D. Vairamuthu, “Context Aware E-Learning using Ontology Framework,” vol. 2, no. 2, pp. 1–7, 2014.
37. A. J. Mbendera, C. Kanjo, and L. Sun, “Towards development of personalised knowledge construction model for e-learning,” Second Int. Conf. Mobile, Hybrid, On-Line Learn., pp. 29–35, 2010.
38. E. K. Kahiigi, L. Ekenberg, H. Hansson, and F. F. Tusubira, “Exploring the e-Learning State of Art,” Electron. J. e-Learning, vol. 6, no. 2, pp. 77–88, 2008.
39. L. Sun, K. Osmanou, and S. a. Williams, “Articulation of Learners Requirements for Personalised Instructional Design in E-Learning Services,” Springer-Verlag Berlin Heidelb., pp. 424–431, 2004.
40. M. Csikszentmihalyi, Flow: The Psychology of Optimal Experience. New York, NY: Harper and Row, 1990.
41. G. B. Moneta, “On the Measurement and Conceptualization of Flow,” in Advances in Flow Research, 2012, p. 231.
42. L. Liao, “A Flow Theory Perspective on Learner Motivation and Behavior in Distance Education,” Distance Educ., vol. 27, no. 1, pp. 45–62, 2006.
43. D. Sampson and P. Zervas, “Context-Aware Adaptive and Personalized Mobile Learning Systems,” in Ubiquitous and Mobile Learning in the Digital Age SE  – 1, D. G. Sampson, P. Isaias, D. Ifenthaler, and J. M. Spector, Eds. Springer New York, 2013, pp. 3–17.
44. D. G. Sampson and P. Zervas, “Ubiquitous and Mobile Learning in the Digital Age,” pp. 3–17, 2013.