Content uploaded by Quintin Siebers
Author content
All content in this area was uploaded by Quintin Siebers
Content may be subject to copyright.
A preview of the PDF is not available
A Context Aware Recommender System for Creativity Support Tools
Abstract and Figures
The development of methods that can enhance the creativity process is becoming a continuous necessity. Through the years several researchers modelled and defined creativity focusing to the psychological aspect of the topic. More recent researchers approach creativity as a computerized process by simulating it within creativity support tools (CST). This article supports that usage of context aware recommender system, in creativity support tools and more specifically, collaborative creativity support tools (CCST) can enhance creativity process. In this work we focus on the development of a context awareness recommender system and look into how such a system can be useful for the creativity process, through preliminary evaluation results in regards to its usefulness and usability.
Figures - uploaded by Quintin Siebers
Author content
All figure content in this area was uploaded by Quintin Siebers
Content may be subject to copyright.
... Concerning the recommender type, six works are classified as CB by their authors (Guan et al., 2017;Koch & Landes, 2014;Serrano, Romero, & Olivas, 2013;Torre & Torsani, 2016;Valentin, Emrich, Lahann, Werth, & Loos, 2015;Wan & Niu, 2018), four proposals are considered CF (Aimeur, Onana, & Saleman, 2007;Bakhshinategh, Spanakis, Zaiane, & Elatia, 2017;Capuano, Gaeta, Ritrovato, & Salerno, 2014;Santos & Boticario, 2008), six researches are KB (Baneres & Conesa, 2017; Emmenegger et al., 2016;Florian & Fabregat, 2011;Harrathi, Touzani, & Braham, 2018;Mao, Shou, Fan, Chen, & Kankanhalli, 2015;Paquette, 2016), one system is treated as UB (Sielis et al., 2011) and the rest are cataloged as hybrid approaches. With regard to the domain, we classify the proposals into: (1) health domain (Khobreh, Ansari, Dornhofer, & Fathi, 2013), (2) communication domain (Guan et al., 2017;Mao et al., 2015;Torre & Torsani, 2016), (3) social media domain (Duran, Chanchi, Arciniegas, & Baldassarri, 2016;Valentin et al., 2015), (4) employment domain (Baneres & Conesa, 2017;Damiani et al., 2015;Isaias, Casaca, & Pifano, 2010;Wang, 2016) and general domain (the rest of proposals). ...
... First criterion indicates whether the recommender system is accessed through a web application (25 of 27), a desktop application (0 of 27) or both possibilities (Duran et al., 2016;Mao et al., 2015). Second criterion details if the recommender system is a stand-alone application (18 of 27), if it is embedded in other system (Capuano et al., 2014;Chavarriaga et al., 2014;Emmenegger et al., 2016;Montuschi et al., 2015;Paquette, 2016;Santos & Boticario, 2008;Sielis et al., 2011;Wang, 2016) or if both values are possible (Duran et al., 2016). The target user is a very relevant criterion since it refers to the person who receives the recommendation. ...
... It provides the user with a survey to assess the recommender system. Six approaches receive an acceptable (Capuano et al., 2014;Florian & Fabregat, 2011), satisfactory (Sielis et al., 2011), good (Baneres & Conesa, 2017;Wan & Niu, 2018) or excellent (Santos & Boticario, 2008) qualification. The remainder proposals do not provide enough information about this criterion. ...
Context
Competences represent an interesting pedagogical support in many processes like diagnosis or recommendation. From these, it is possible to infer information about the progress of the student to provide help targeted both, trainers who must make adaptive tutoring decisions for each learner, and students to detect and correct their learning weaknesses. For the correct development of any of these tasks, it is important to have a suitable student model that allows the representation of the most significant information possible about the student. Additionally, it would be very advantageous for this modeling to incorporate mechanisms from which it would be possible to infer more information about the student’s state of knowledge.
Objective
To facilitate this goal, in this paper a new approach to develop an adaptive competence-based recommender system is proposed.
Method
We present a methodological development guide as well as a set of ontological and non-ontological resources to develop and adapt the prototype of the proposed recommender system.
Results
A modular flexible ontology network previously built for this purpose has been extended, which is responsible for recording the instructional design and student information. Furthermore, we describe a case study based on a first aid learning experience to assess the prototype with the proposed methodology.
Conclusions
We highlight the relevance of flexibility and adaptability in learning modeling and recommendation processes. In order to promote improvement in the personalized learning of students, we present a Recommender System prototype taking advantages of ontologies, with a methodological guide, a broad taxonomy of recommendation criteria and the nature of competences. Future lines of research lines, including a more comprehensive evaluation of the system, will allow us to demonstrate in depth its adaptability according to the characteristics of the student, flexibility and extensibility for its integration in various environments and domains.
... • UB recommender systems are uncommon in all areas and in this survey, only one proposal was found (Sielis et al. 2011). ...
... Implementation of recommender systems.an expert user(Aimeur, Onana, and Saleman 2007;Sielis et al. 2011;Emmenegger et al. 2016).Second category contains the works that recommend any kind of resource. In this latter group, there are proposals recommending learning objects(Rodríguez, Ovalle, and Duque 2015), learning materials(Khobreh et al. 2013), general resources(Paquette 2016;Sielis et al. 2011; Valentin et al. 475 2015) or others(Torre and Torsani 2016;Guan et al. 2017). ...
... expert user(Aimeur, Onana, and Saleman 2007;Sielis et al. 2011;Emmenegger et al. 2016).Second category contains the works that recommend any kind of resource. In this latter group, there are proposals recommending learning objects(Rodríguez, Ovalle, and Duque 2015), learning materials(Khobreh et al. 2013), general resources(Paquette 2016;Sielis et al. 2011; Valentin et al. 475 2015) or others(Torre and Torsani 2016;Guan et al. 2017). The third cluster includes the course recommendation's system from student's information, mainly, profile and competences(Koch and Landes 2014;Montuschi et al. 2015;Baneres and Conesa 2017;Bakhshinategh et al. 2017). ...
Competence-based learning is increasingly widespread in many institutions since it provides flexibility, facilitates the self-learning and brings the academic and professional worlds closer together. Thus, the competence-based recommender systems emerged taking the advantages of competences to offer suggestions (performance of a learning experience, assistance of an expert or recommendation of a learning resource) to the user (learner or instructor). The objective of this work is to conduct a new Systematic Literature Review (SLR) concerning competence-based recommender systems to analyse in relation to their nature and assessment of competences an others key factors that provide more flexible and exhaustive recommendations. To do so, a SLR research methodology was followed in which 25 competence-based recommender systems related to learning or instruction environments were classified according to multiple criteria. We evaluate the role of competences in these proposals and enumerate the emerging challenges. Also a critical analysis of current proposals is carried out to determine their strengths and weakness. Finally, future research paths to be explored are grouped around two main axes closely interlinked; first about the typical challenges related to recommender systems and second, concerning ambitious emerging challenges.
... First, the recommendation mechanism could be improved in different ways. This study, as well as other related works such as recommenders for creativity [31] and scientific writing [32], relies on topic modeling and bag-of-word models to find recommendations. Encoding text using attention-based models [33], such as BERT [34], have been shown to perform well on various natural language processing tasks, including semantic sentence similarity for conversation data [28]. ...
Background
The working environment of a suicide prevention helpline requires high emotional and cognitive awareness from chat counselors. A shared opinion among counselors is that as a chat conversation becomes more difficult, it takes more effort and a longer amount of time to compose a response, which, in turn, can lead to writer’s block.
Objective
This study evaluates and then designs supportive technology to determine if a support system that provides inspiration can help counselors resolve writer’s block when they encounter difficult situations in chats with help-seekers.
Methods
A content-based recommender system with sentence embedding was used to search a chat corpus for similar chat situations. The system showed a counselor the most similar parts of former chat conversations so that the counselor would be able to use approaches previously taken by their colleagues as inspiration. In a within-subject experiment, counselors’ chat replies when confronted with a difficult situation were analyzed to determine if experts could see a noticeable difference in chat replies that were obtained in 3 conditions: (1) with the help of the support system, (2) with written advice from a senior counselor, or (3) when receiving no help. In addition, the system’s utility and usability were measured, and the validity of the algorithm was examined.
Results
A total of 24 counselors used a prototype of the support system; the results showed that, by reading chat replies, experts were able to significantly predict if counselors had received help from the support system or from a senior counselor (P=.004). Counselors scored the information they received from a senior counselor (M=1.46, SD 1.91) as significantly more helpful than the information received from the support system or when no help was given at all (M=–0.21, SD 2.26). Finally, compared with randomly selected former chat conversations, counselors rated the ones identified by the content-based recommendation system as significantly more similar to their current chats (β=.30, P<.001).
Conclusions
Support given to counselors influenced how they responded in difficult conversations. However, the higher utility scores given for the advice from senior counselors seem to indicate that specific actionable instructions are preferred. We expect that these findings will be beneficial for developing a system that can use similar chat situations to generate advice in a descriptive style, hence helping counselors through writer’s block.
... First, the recommendation mechanism could be improved in different ways. This study, as well as other related works such as recommenders for creativity [31] and scientific writing [32], relies on topic modeling and bag-of-word models to find recommendations. Encoding text using attention-based models [33], such as BERT [34], have been shown to perform well on various natural language processing tasks, including semantic sentence similarity for conversation data [28]. ...
BACKGROUND
The working environment of a suicide prevention helpline requires high emotional and cognitive awareness from chat counselors. A shared opinion among counselors is that as a chat conversation becomes more difficult, it takes more effort and a longer amount of time to compose a response, which, in turn, can lead to writer’s block.
OBJECTIVE
This study evaluates and then designs supportive technology to determine if a support system that provides inspiration can help counselors resolve writer’s block when they encounter difficult situations in chats with help-seekers.
METHODS
A content-based recommender system with sentence embedding was used to search a chat corpus for similar chat situations. The system showed a counselor the most similar parts of former chat conversations so that the counselor would be able to use approaches previously taken by their colleagues as inspiration. In a within-subject experiment, counselors’ chat replies when confronted with a difficult situation were analyzed to determine if experts could see a noticeable difference in chat replies that were obtained in 3 conditions: (1) with the help of the support system, (2) with written advice from a senior counselor, or (3) when receiving no help. In addition, the system’s utility and usability were measured, and the validity of the algorithm was examined.
RESULTS
A total of 24 counselors used a prototype of the support system; the results showed that, by reading chat replies, experts were able to significantly predict if counselors had received help from the support system or from a senior counselor ( P =.004). Counselors scored the information they received from a senior counselor (M=1.46, SD 1.91) as significantly more helpful than the information received from the support system or when no help was given at all (M=–0.21, SD 2.26). Finally, compared with randomly selected former chat conversations, counselors rated the ones identified by the content-based recommendation system as significantly more similar to their current chats (β=.30, P <.001).
CONCLUSIONS
Support given to counselors influenced how they responded in difficult conversations. However, the higher utility scores given for the advice from senior counselors seem to indicate that specific actionable instructions are preferred. We expect that these findings will be beneficial for developing a system that can use similar chat situations to generate advice in a descriptive style, hence helping counselors through writer’s block.
... Based on these components, tutors can manually define and parameterize recommendation rules, which will only trigger a recommendation if conditions regarding categories, inputs and settings are met. Thus, LIME is a tutor/teacher crafted, rule-based recommender system for cloud learning environments (SPOCs or MOOCs), which contrasts with other approaches (Lenoy et al., 2013;Sielis et al., 2011). ...
Formal and informal learning settings overlap quite often. Since the arrival and consolidation of 24/7 Internet services, contents, authoring tools and free storage, informal settings have become a powerful resource for learning. The combination of both scenarios is a reality that must be tackled: one with regular academic programmes, and the other one with mostly unstructured learning and information unit. However, successful engagement between them is not that common. Regular academic programmes deployed with learning management systems (LMS) keep a track of user activity and user performance. Sometimes these inputs are stored and processed to provide some useful feedback in the form of recommendation or advice to the very same user. However, there is usually no tracking of user activity on social media, external fora or other online services. This lack of awareness about user activity outside the LMS prevents the educational methodology or the pedagogical model from considering all the available information. It also restricts the analysis to just one area: official education. To this extent, MOOCs have become a crucial part of combined educational models that move between formal and informal settings, where they play a key role in the learning path of every user. However, this does not seem enough. The current educational landscape requires disruption to boost the learning-teaching process. We call this disruption transgenic learning. It is based on the user’s behaviour and interactions, along with efficient monitoring and personalised counselling by a tutor. It is focused on improving the user’s awareness of their real academic status and their related performance. In this book chapter, we discuss the specific implementation of a personalised eLearning model for restricted social networks and learning management systems, called LIME, which supports this approach. LIME is focused on massive enrolment and large datasets. We also present a framework and software prototype that implements the model (iLIME). Lastly, we show a successful practical case study that makes use of iLIME integrated with the Sakai CLE LMS. This case study depicts the technical issues, challenges and solutions involved in a successful deployment of the LIME model into a real university scenario.
Recommender systems are decision support systems helping users to identify one or more items (solutions) that fit their wishes and needs. The most frequent application of recommender systems nowadays is to propose items to individual users. However, there are many scenarios where a group of users should receive a recommendation. For example, think of a group decision regarding the next holiday destination or a group decision regarding a restaurant to visit for a joint dinner. The goal of this book is to provide an introduction to group recommender systems, i.e., recommender systems that determine recommendations for groups. In this chapter, we provide an introduction to basic types of recommendation algorithms for individual users and characterize related decision tasks. This introduction serves as a basis for the introduction of group recommendation algorithms in Chap. 2.
Team creativity can be strongly affected by communication. With the trend of online collaboration, the design of communication tools increasingly affects different creative processes of teams working online. To clarify this effect, this study took the perspective of affordances and reviewed 54 papers to map the needs of teams, the design of communication tools, and the effects of tools on creativity in different creative stages. First, we summarized teams’ requirements for communication tools in different stages. Second, we identified key affordances with relevant features. Third, we discussed how these affordances could affect communication and team creativity in different stages and built a mapping of creative teams’ needs and communication tools. The mapping can provide insights for both designing communication tools for creative teams and expanding the current team creativity theories to fit the online context and new communication technologies.KeywordsCommunicationCommunication ToolsTeam CreativityCreative ProcessOnline Collaboration
Online whiteboards are becoming a popular way to facilitate collaborative design work, providing a free-form environment to curate ideas. However, as templates are increasingly being used to scaffold contributions from non-experts designers, it is crucial to understand their impact on the creative process. In this paper, we present the results from a study with 114 students in a large introductory design course. Our results confirm prior findings that templates benefit students by providing a starting point, a shared process, and the ability to access their own work from previous steps. While prior research has criticized templates for being too rigid, we discovered that using templates within a free-form environment resulted in visual patterns of free-form curation where concepts were spatially organized, clustered, color-coded, and connected using arrows and lines. We introduce the concept of 'Free-form Templates' to illustrate how templates and free-form curation can be synergistic.
In the Artificial Intelligence (AI) field, and particularly within the area of Machine Learning (ML), recommender systems have attracted significant research attention. These systems attempt to alleviate the increasing information overload that users can experience in the current Big Data era, by providing personalized recommendations of items that they may find relevant. Besides, given the importance of mobile computing, these systems have evolved to consider also the dynamic context of the mobile users (location, time, weather conditions, etc.) to offer them more appropriate suggestions and information while on the move.
In this paper, we provide an extensive survey of recent advances towards intelligent mobile Context-Aware Recommender Systems (mobile CARS) from an information management perspective, with an emphasis on mobile computing and AI techniques, along with an analysis of existing research gaps and future research directions. We focus on approaches that go beyond just considering the location of the user and exploit also other context information. In this study, we have identified that deep learning approaches are promising artificial intelligence models for mobile CARS. Additionally, in a near future, we expect a higher prominence of push-based recommendation solutions where at least part of the recommendation engine could be executed in the mobile devices, which could share data and tasks in a distributed way.
Recommender systems are decision support systems helping users to identify one or more items (solutions) that fit their wishes and needs. The most frequent application of recommender systems nowadays is to propose items to individual users. However, there are many scenarios where a group of users should receive a recommendation. For example, think of a group decision regarding the next holiday destination or a group decision regarding a restaurant to visit for a joint dinner. The goal of this book is to provide an introduction to group recommender systems, i.e., recommender systems that determine recommendations for groups. In this chapter, we provide an introduction to basic types of recommendation algorithms for individual users and characterize related decision tasks. This introduction serves as a basis for the introduction of group recommendation algorithms in Chap. 2.
Recommender systems represent user preferences for the purpose of suggesting items to purchase or examine. They have become fundamental applications in electronic commerce and information access, providing suggestions that effectively prune large information spaces so that users are directed toward those items that best meet their needs and preferences. A variety of techniques have been proposed for performing recommendation, including content-based, collaborative, knowledge-based and other techniques. To improve performance, these methods have sometimes been combined in hybrid recommenders. This paper surveys the landscape of actual and possible hybrid recommenders, and introduces a novel hybrid, EntreeC, a system that combines knowledge-based recommendation and collaborative filtering to recommend restaurants. Further, we show that semantic ratings obtained from the knowledge-based part of the system enhance the effectiveness of collaborative filtering.
Interactive systems have been the dominant computing paradigm over recent years. This paradigm is characterized by the fact that human user and the system communicate and interact explicitly using different modalities. However to come closer to visions of Ambient Intelligence, Calm Computing, Disappearing Computing, and Ubiquitous Computing new forms of interaction are required. Observing humans interacting with each other and new possibilities given by emerging technologies indicate that a new interaction model is needed. In this chapter we present the concept of implicit human computer interaction (iHCI) that takes the users context into account when creating new user interfaces for ambient intelligence. Beyond the model examples are given, application areas are described and basic implementation issues are discussed. Our research leads to a more general discussion on disappearing and invisible user interfaces. In the invisibility dilemma we explain that in many application areas there may be an inherent conflict. The transparent user interface and the added value gained by introducing technology are often opposing goals, especially combined with users that are creative in appropriating their tools.
The question of whether creativity is domain general or domain specific is one of the most enduring controversies in the field. Strong opinions are regularly shared in support of both positions, which has led to a polarization of the debate. The dominant perspective currently appears to be that of domain specificity, probably due in large part to the pendulum swing toward situated cognition in the social sciences in general. However, considerable evidence supports the idea that creativity has both specific and general components, and that the level of specificity-generality changes with the social context and as one develops through childhood into adulthood. The authors present their position that creativity is a developmental construct that can be viewed as exhibiting both domain-specific and domain-general characteristics. Furthermore, they argue that from an educational perspective, these distinctions simply are not very important. In this chapter, they present their conception of creativity, describe why creativity can (and should) be viewed as both context-free and context-dependent, and review the theoretical, empirical, and educational implications of this conceptualization using a preliminary model drawn from their analysis. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
The paper presents the work of the Elena project. The contribution of this paper is twofold: The first part of the paper elaborates on the notion of learning services. Learning services are differentiated from learning objects and put in the context of web services. A model for learning services is introduced and current limitations of learning object metadata standards are addressed. The second part of the paper sketches a mediation infrastructure for learning ser-vices, called Smart Space for Learning. The architecture of Smart Spaces for Learning is based on P2P and semantic web technolo-gies.
The vision of ubiquitous computing is about numerous devices embedded in our every-day environment, designed to serve humans
in a non-obtrusive manner while minimizing the required user attention. These devices are expected to seamlessly monitor context
changes and adapt their behavior and functionality to maximize the user benefit. However, designing the self-adaptive logic
of such systems is far from trivial. This paper discusses a utility function-based approach for specifying the adaptive behavior
of component-based context-aware systems. Although this approach allows for completely autonomous self-adaptive behavior,
it also leverages potential user feedback by adjusting and optimizing its behavior. It is argued that this approach provides
significant improvement to the adaptive behavior of a system while maintaining the required user attention to a minimum. At
the same time, it keeps the complexity involved in the development of such context-aware, self-adaptive applications to a
reasonably low level.
This paper presents an overview of the field of recommender systems and describes the current generation of recommendation methods that are usually classified into the following three main categories: content-based, collaborative, and hybrid recommendation approaches. This paper also describes various limitations of current recommendation methods and discusses possible extensions that can improve recommendation capabilities and make recommender systems applicable to an even broader range of applications. These extensions include, among others, an improvement of understanding of users and items, incorporation of the contextual information into the recommendation process, support for multicriteria ratings, and a provision of more flexible and less intrusive types of recommendations.
To support users in performing their tasks, applications need a better understanding of the current situation they are being used in. This chapter gives an overview of how knowledge of the current context, that is, information characterizing the situation, can be represented and how this knowledge can be used for enhancing applications. We discuss what is actually meant by "context" and "context-aware" applications. Further, we describe what has to be considered when building a context-aware application. We thereby focus on the representation of context information and how to deal with its unreliable nature. This chapter should sensitize the reader to the difficulties of using context information and give guidelines on how to build an application that benefits from knowing its current context.
A new era of e-learning is on the horizon, hundreds of Learning Contents are created and more and more people begin to acquire acknowledge thru e-learning. The traditional teaching method is already showing its limitations that students from different backgrounds are still given the same contents at the same time, and they may only interest in part of a whole learning content. In this paper, we propose a novel way to organize learning contents into small "atomic" units called Learning Objects so that they could be used and reused effectively. The Learning Objects together with their ontology are systemized into knowledge base. An intelligent recommendation mechanism based on sequencing rules is then introduced with detail, where the rules are formed from the knowledge base and competency gap analysis. Finally we establish a test knowledge base system, using and extending the ontology editor Protégé-2000 and its Protégé Axiom Language.. She received her master's degree in Electrical Engineering (Circuits and System) from Shanghai Jiaotong University in 2000 and now is a PhD candidate in Computer Science and Technology. She is also a member of the content based working group of the Chinese e-Learning Technology Standardization Committee (CELTSC). Her research interests lie in the area of computer networking and intelligent e-learning system. She has published papers in several conference proceedings and journals.