15 research outputs found
Issues and Considerations regarding Sharable Data Sets for Recommender Systems in Technology Enhanced Learning
No full textDrachsler, H., Bogers, T., Vuorikari, R., Verbert, K., Duval, E., Manouselis, N., Beham, G., Lindstaedt, S., Stern, H., Friedrich, M., & Wolpers, M. (2010, 28 September). Issues and Considerations regarding Sharable Data Sets for Recommender Systems in Technology Enhanced Learning. Presentation at the 1st Workshop Recommender Systems in Technology Enhanced Learning (RecSysTEL) in conjunction with 5th European Conference on Technology Enhanced Learning (EC-TEL 2010): Sustaining TEL: From Innovation to Learning and Practice, Barcelona, Spain.The presentation is based on the positioning paper of the dataTEL Theme Team of the STELLAR Network of Excellence (http://www.teleurope.eu/pg/groups/9405/datatel/) that addresses the lack of educational data sets in TEL and present ideas to overcome this situation. The accompanying paper: Issues and Considerations regarding Sharable Data Sets for Recommender Systems in Technology Enhanced Learning, can be found at http://www.sciencedirect.com/science/journal/18770509 and a pre-print is available in our Dspace repository and at scribd.
The presentation starts with a description of the current situation where almost none educational data sets are publicly available. This is a strange situation as plenty of data is saved on a daily base in LMS like Moodle, Blackboard. In other domains like e-commerce it is a common practice to use publicly available data sets from different application environments (e.g. Yahoo, MovieLens) in order to evaluate algorithms and create new data products. These data sets are for instance used as benchmarks to develop new recommendation algorithms and compare them to other algorithms in certain settings.
Recommender systems are also increasingly applied in Technology Enhanced Learning field but it is still an application area that lacks such publicly available data sets. Although there is a lot of research conducted on recommender systems in TEL, they lack data sets that would allow the experimental evaluation of the performance of different recommendation algorithms using comparable, interoperable, and reusable data sets. This leads to awkward experimentation and testing such as using data sets from movies in order to evaluate educational recommendation algorithms.Stella
Collecting copper plates between Venice and Rome in the seventeenth century. Cardinal Alessandro Orsini, the Old Masters and the sciences
No full textCardinal Alessandro Orsini is principally characterized by his passion for science and interest in the arts. New research has made it possible to reconstruct his biography and the dynamics that led to the formation of his art collection. Uniquely in Rome, it comprised numerous copper plates identified by the author as works by Albrecht Dürer, Lucas van Leyden and Hans Sebald Beham. The Cardinal was among the main patrons of the Stamperia Ducale at Bracciano; the author suggests that he purchased the copper plates in Venice from the merchant Daniel Nijs. This has made it possible to trace a hitherto unknown story involving the dissemination of art works, made possible by Orsini’s constant journeys between Rome, Bracciano and the cities of Northern Europe, which allowed him to establish relationships with the period’s most important thinkers, develop open-minded intellectual views and increase his own collection with a preference for Venetian painting
Review of: Tina Schmid: Generation, Geschlecht und Wohlfahrtsstaat. Intergenerationelle Unterstützung in Europa. Wiesbaden: Springer VS 2014.
No full textAuf der Basis der SHARE-Studie (Survey of Health, Ageing and Retirement in Europe) untersucht Tina Schmid Geschlechterunterschiede in den Unterstützungsleistungen zwischen den Generationen. Dabei nimmt sie Leistungen von Eltern an erwachsene Kinder und gleichzeitig von erwachsenen Töchtern und Söhnen an ihre Eltern in den Blick. Der innovative Charakter des Buches liegt neben dieser simultanen Betrachtung von Unterstützungsleistungen in beide Richtungen darin, dass die Autorin durch die Verknüpfung von Fragestellungen und Konzeptionen der Generationenforschung mit jenen der Geschlechtersoziologie und der komparativen Wohlfahrtsstaatenforschung zu differenzierten Erkenntnissen und zu einer Perspektivenerweiterung hinsichtlich zeitlicher Unterstützungsleistungen beiträgt.Based on the SHARE-study (Survey of Health, Ageing and Retirement in Europe), Tina Schmid examines gender differences in the benefits between the generations. In doing so, she concurrently takes into account benefits from parents to their adult children as well as benefits from daughters and sons to their parents. The book’s innovative character is due to the author’s simultaneous analysis of benefits in both directions as well as the fact that, by linking questions and concepts from genealogy with those from gender sociology and the comparative study of welfare states, the author provides complex findings and contributes to the expansion of perspectives regarding temporal benefits
Seeing What the System Thinks You Know: Visualizing Evidence in an Open Learner Model
No full textUser knowledge levels in adaptive learning systems can be assessed based on user interactions that are interpreted as Knowledge Indicating Events (KIE). Such an approach makes complex inferences that may be hard to understand for users, and that are not necessarily accurate. We present MyExperiences, an open learner model designed for showing the users the inferences about them, as well as the underlying data. MyExperiences is one of the first open learner models based on tree maps. It constitutes an example of how research into open learner models and information visualization can be combined in an innovative way
Fitness Landscape Analysis and Algorithm Selection for Assignment Problems
No full textOptimierungsalgorithmen finden vielfach Anwendung bei Planungs- und Steuerungsaufgaben. Im Betrieb von Lastkraftwagen, Shiffen, Kränen, Produktionslinien, Spitälern oder Lagerhäuser sind Entscheidungen zu treffen, die die Effizienz, Handlungsmöglichkeiten und Kosten beeinflussen. Deterministische und stochastische Modelle formalisieren diese Entscheidungssituationen und deren Ziele. Exakte und heuristische Ansätze werden zur effizienten Lösung solcher Modelle eingesetzt und entsprechende Entscheidungen werden aus deren Lösungen abgeleitet. Eine Vielzahl unterschiedlicher Algorithmen zur Lösungsfindung sind verfügbar, übertreffen sich jedoch gegenseitig bei unterschiedlichen Problemen. Es ist daher entscheidend den richtigen Algorithmus auszuwählen. Ziel dieser Arbeit ist es, Methoden die das sogenannte Algorithmus Auswahlproblem lösen, zu entwickeln und zu evaluieren. Wir werden verbesserte Methoden auf Basis bestehender Algorithmen entwicklen und in der Anwendung auf bekannte Modelle untersuchen. In dieser Hinsicht werden Forschungs- und Entwicklungstätigkeiten aus drei Bereichen zusammengeführt. Zuerst wird das Gebiet der Fitnesslandschaftsanalyse vorgestellt und wie damit Modelleinstanzen durch Merkmale beschrieben werden. Nach einem Überblick über Landschaftsaspekte und deren Einflüsse auf das Leistungsverhalten von Algorithmen, werden neue Methoden und Merkmale vorgestellt und analysiert. Aufbauend darauf werden Ideen für neue Algorithmen diskutiert. Im zweiten Teil der Arbeit wird das Algorithmus Auswahlproblem in zwei Fallstudien gelöst. Mehrere tausend Versuche wurden hierfür auf einer breiten Palette von Modellinstanzen durchgeführt. Aus dem beobachteten Konvergenzverhalten vieler Algorithmen wird in den beiden Fallbeispielen ein klassifikationsbasierter Ansatz zur Lösung des Auswahlproblems evaluiert. Die Resultate aus den Versuchen werden vergleichend gegenübergestellt und Zusammenhänge mit Landschaftseigenschaften untersucht. Wir werden zeigen, dass ein kombinierter Ansatz in der Lage ist einzelne Algorithmen zu übertreffen. Im dritten Teil der Arbeit wird, anhand einem allgemeinen Vorgehensmodell, diskutiert wie sich dies auf andere Modelle und Domänen übertragen lässt. Dabei werden Entscheidungen beschrieben, die für die Überführung einer echten Entscheidungssituation in ein wissenschaftliches Modell notwendig sind. Zum Schluß werden Softwaresysteme vorgestellt, die solche Entscheidungen unterstützen und die vom Autor während seines Studiums erstellt wurden.The field of optimization has a wide range of applications in the real world. The operation of trucks, ships, cranes, production lines, hospitals, or warehouses results in many decision situations that affect efficiencies, business capabilities, and costs. Deterministic and stochastic models formalize the decision situations as well as their objectives. Exact and heuristic techniques are employed to solve these models efficiently. The resulting solutions are then used to derive the respective decisions. A large range of such solving algorithms are available, but which outperform each other in different situations. It is thus a major problem to decide which algorithm instance should be applied to solve a specific model. It is the goal of this thesis to devise and evaluate methods that are able to solve this so called algorithm selection problem. We will evaluate whether an improved solving algorithm can be found that reuses existing approaches. The goal is to evaluate such an algorithm when applied to well-known models and evaluate its performance. In this light, research and development efforts from three domains are described and studied. First, fitness landscape analysis is a method applicable to characterize models and their instances through a set of features. We will provide an overview on high-level characteristics that influence the performance of solving algorithms. In this thesis a new method termed directed walk and new features are devised, different variants thereof are analyzed, and new ideas for algorithms are given. Second, the algorithm selection problem is solved in two case studies. Several thousand runs have been performed on a wide range of different benchmark instances. Thereby, the performance of a range of algorithms is recorded. In these two studies we will evaluate a classification-based approach to the algorithm selection problem. Thereby, we will compare the performances of algorithms and relate those to the landscape characteristics. We will show that a combined approach performs better than an individual algorithm. Third, the general research model of operations research is introduced and it is discussed how such results can be transferred to other models and domains. Thereby, we discuss the steps involved to move from a real-world problem to a scientific model and further to a solution through an efficient solver. Finally, we introduce software systems that support such tasks and which have been created by the author in the course of his studies.submitted by Andreas BehamDissertation Universität Linz 201
Teaching and participatory media
No full textHow can diversity of teaching practice be appreciated? Perhaps to appreciate diversity we should consider where we have arrived at a consensus, and hence conformity. It is normally expected that mobile phones should be switched off during lectures. For this talk the speaker will request that all mobiles be turned on (though set to silent). As this is against most orthodox teaching practice it could reasonably be concluded that the speaker is not an authority fit to lecture on the subject. For this reason the speaker will instruct the audience not to learn anything during the talk. Notes should not be taken. Instead listeners will be encouraged to send text messages. These messages will be in the form of single words, short phrases or multiple choice answers to questions (similar to classroom voting systems).
It was recently reported that the UK government is considering introducing Twitter to its primary school curriculum (Guardian, 2009). This paper will evaluate ways in which new participatory media are impacting education. A free tool created by the author will be used during the oral presentation of this paper to show how these technologies allow increased participation in learning scenarios. One of the significant aspects of teaching strategies that are utilising text messaging and micro-blogging is that they are allow for teachers to delegate control to learners whilst keeping the format of learner participation sufficiently constrained for use with groups. This can be seen as an application of education research on the concept of Control and how its negotiation and subsequent delegation can be used to increase the degree of autonomy for a learner within a learning trajectory (Moore, 1993; Dron, J. 2007).
References:
Dron, J. (2007) Control and constraint in e-learning: Choosing when to choose. Idea Group Publishing.
Moore, M. G. (1993) 2 Theory of transactional distance. Theoretical principles of distance education. pp22.
The Guardian (2009) Pupils to study Twitter and blogs in primary schools shake. 25 March 2009, Retrieved from http://www.guardian.co.uk/education/2009/mar/25/primary-schools-twitter-curriculu
Sebald Beham’s Peasant Dances in Historical, Social and Cultural Context
No full textThe German so-called Little Masters (Kleinmeister) were painter-engravers, active in the first half of the 16th century. They did
not form any group or belong to any school, but were direct continuators of Albrecht Dürer’s printmaking. Depending on the author,
the term Little Masters has been applied to various artists from this generation, the most significant ones being the three artists from
Nuremberg: (Hans) Sebald (1500-1550), Barthel (1502-1540) Beham and Georg Pencz (ca. 1500-1550), as well as Albrecht Altdorfer
(1480-1538) and Heinrich Aldegrever (1502-ca. 1555), and the lesser known Hans Brosamer (ca. 1490-1552), Jacob Binck (1494/1500-after 1569) and an anonymous Master IB. The article focuses on Sebald Beham whose biography and works have been repeatedly studied in German and English literature,but only mentioned in Polish publications. As far as his education is concerned, it is possible, but not documented, that he was trained by his brother Barthel and subsequently by Dürer. His life has often been regarded as a rebellious one due to his troubles with the authorities; firstly, when he confessed, together with his brother and Georg Pencz, his disbelief in the holy sacraments and his disrespect for the town council, and secondly, when his book on the proportions of the horse was published before Dürer’s Art of Measurement. In the first case, he was jailed and expelled from Nuremberg for just under a year, whereas after the second scandal he left the city for few months. This prolific artist’s oeuvre encompasses about 250 copperplates, 18 etchings and over 1500 woodcuts, including book illustrations. He was the author of simple and small woodcuts, extremely detailed miniature works in metal created with thin and delicate lines, as well as large woodcuts used as wallpapers. Moreover, the subjects of his prints were varied: Biblical (mostly the Old Testament), mythological, historical (ancient history in particular) and genre scenes (mainly peasants), images of saints, portraits, allegorical representations, coats of arms, decorative and ornamental motifs. My paper deals with Beham’s prints depicting peasant dances and festivals, especially the three series of engravings created in Frankfurt am Main between 1535-1540: Peasant Festival (1537), Peasant Wedding Procession (ca. 1538–1540) and Peasant Festival or The Twelve Months (1546–1547). Although it was Dürer who introduced this genre motif into the graphic arts and inspired his followers, the creation and development of this specific type of representation is ascribed to the Beham brothers. Sebald’s large woodcuts and miniature engravings depicted the villagers’ customs in different ways; taking the audience’s needs into consideration, he developed, changed and re-edited this motif in his prints till the second half of the 16th century, when it began to function as a separate subject. The development of the peasant motif in graphic arts at that time resulted from several factors. First of all, it occurred during a period of interest in the way of life and feasts of this social group. Studies on the past and the first attempts to create a national identity provide an explanation for this fascination, as does the changing atmosphere of Nuremberg secondary to social and religious transformations. The new Lutheran doctrine condemned church holidays as being more secular than religious in character. Kermis, the most popular festival, which celebrated a church’s anniversary or its saint’s day, was often an occasion that provoked numerous sins: blasphemy, drunkenness, anger, lust or adultery, as well as quarrels, brawls and even murders. As a result, the aforementioned feasting became a worrying issue for the town council, which gradually began to cancel some of the festivals; the authorities tried to make both the upper and lower classes aware of the destructive effects of village holidays. Taking the moralistic and didactic character of Beham’s graphic works into consideration, one can assume that the artist expressed recognition as well as critique of this kind of feast. Last but not least, the Peasants’ War also accounted for rising interest in rural life, although it cannot be regarded as a key factor. While most researchers have tried to find only one right interpretation as far as both large woodcuts and the series of miniature engravings depicting peasants are concerned, it seems to me that one should study these works in a wider context, taking their varied contents, symbols and meanings into consideration. These do not have to exclude one another; on the contrary, they can complete each other and help us to understand Beham’s prints better in their historical, social and cultural context. Such an approach gives rise to new possibilities of interpretation
Role resources and work-family enrichment: The role of work engagement
No full textThe majority of work-family research has focused on negative spillover between demands and outcomes and between the work and family domains (e.g., work-family conflict; see review by Eby, Casper, Lockwood, Bordeaux, & Brinley, 2005). The theory that guided this research was in most cases role stress theory (Greenhaus & Beutell, 1985) or the role scarcity hypothesis (Edwards & Rothbard, 2000). However, according to spillover theory, work-related activities and satisfaction also affect non-work performance, and vice versa. Recently, in line with the positive psychology movement (Seligman & Csikszentmihalyi, 2000), work-family interaction research has also included concepts of positive spillover (Bakker & Schaufeli, 2008; Grzywacz & Marks, 2000). This emerging focus supplements the dominant conflict perspective by identifying new ways of cultivating human resource strength
High performance latent dirichlet allocation for text mining
No full textThis thesis was submitted for the degree of Doctor of Philosophy and awarded by Brunel University.Latent Dirichlet Allocation (LDA), a total probability generative model, is a three-tier Bayesian model. LDA computes the latent topic structure of the data and obtains the significant information of documents. However, traditional LDA has several limitations in practical applications. LDA cannot be directly used in classification because it is a non-supervised learning model. It needs to be embedded into appropriate classification algorithms. LDA is a generative model as it normally generates the latent topics in the categories where the target documents do not belong to, producing the deviation in computation and reducing the classification accuracy. The number of topics in LDA influences the learning process of model parameters greatly. Noise samples in the training data also affect the final text classification result. And, the quality of LDA based classifiers depends on the quality of the training samples to a great extent. Although parallel LDA algorithms are proposed to deal with huge amounts of data, balancing computing loads in a computer cluster poses another challenge. This thesis presents a text classification method which combines the LDA model and Support Vector Machine (SVM) classification algorithm for an improved accuracy in classification when reducing the dimension of datasets. Based on Density-Based Spatial Clustering of Applications with Noise (DBSCAN), the algorithm automatically optimizes the number of topics to be selected which reduces the number of iterations in computation. Furthermore, this thesis presents a noise data reduction scheme to process noise data. When the noise ratio is large in the training data set, the noise reduction scheme can always produce a high level of accuracy in classification. Finally, the thesis parallelizes LDA using the MapReduce model which is the de facto computing standard in supporting data intensive applications. A genetic algorithm based load balancing algorithm is designed to balance the workloads among computers in a heterogeneous MapReduce cluster where the computers have a variety of computing resources in terms of CPU speed, memory space and hard disk space
