lnu.sePublications
Change search
Refine search result
1 - 19 of 19
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rows per page
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sort
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
  • Standard (Relevance)
  • Author A-Ö
  • Author Ö-A
  • Title A-Ö
  • Title Ö-A
  • Publication type A-Ö
  • Publication type Ö-A
  • Issued (Oldest first)
  • Issued (Newest first)
  • Created (Oldest first)
  • Created (Newest first)
  • Last updated (Oldest first)
  • Last updated (Newest first)
  • Disputation date (earliest first)
  • Disputation date (latest first)
Select
The maximal number of hits you can export is 250. When you want to export more records please use the Create feeds function.
  • 1.
    Imran, Ali Shariq
    et al.
    Norwegian University of Science and Technology (NTNU), Norway.
    Dalipi, Fisnik
    University of South-Eastern Norway, Norway.
    Kastrati, Zenun
    Linnaeus University, Faculty of Technology, Department of computer science and media technology (CM).
    Predicting Student Dropout in a MOOC: An Evaluation of a Deep Neural Network Model2019In: Proceedings of the 2019 5th International Conference on Computing and Artificial Intelligence, ACM Publications, 2019, p. 190-195Conference paper (Refereed)
    Abstract [en]

    Massive Open Online Courses (MOOCs) have transformed the way educational institutions deliver high-quality educational material to the onsite and distance learners across the globe. As a result, a new paradigm shifts as to how learners acquire and benefit from the wealth of knowledge provided by a MOOC at their doorstep nowadays in contrast to the brick and mortar settings is visible. Learners are therefore showing a profound interest in the MOOCs offered by top universities and industry giants. They have also attracted a vast number of students from far-flung areas of the world. The massive number of registered students in MOOCs, however, pose one major challenge, i.e., 'the dropouts'. Course planners and content providers are struggling to retain the registered students, which give rise to a new research agenda focusing on predicting and explaining student dropout and low completion rates in a MOOC. Machine learning techniques utilizing deep learning approaches can efficiently predict the potential dropouts and can raise an alert well before time. In this paper, we have focused our study on the application of feed-forward deep neural network architectures to address this problem. Our model achieves not only high accuracy, but also low false negative rate while predicting dropouts on the MOOC data. Moreover, we also provide an in-depth comparison of the proposed architectures concerning precision, recall, and F1 measure.

  • 2. Imran, Ali Shariq
    et al.
    Kastrati, Zenun
    Pedagogical Document Classification and Organization Using Domain Ontology2016In: Learning and Collaboration Technologies, Springer International Publishing , 2016, p. 499-509Conference paper (Refereed)
    Abstract [en]

    One of the challenges faced by today’s web is the abundance of unstructured and unorganized information available on the Internet in form of educational documents, lecture notes, presentation slides, and multimedia recordings. Accessing and retrieving the massive amount of such resources are not an easy task, especially educational resources of pedagogical nature. Much of the pedagogical content available on Internet comes from blogs, wikis, posts with little or no metadata, that suffer from the same dilemma. The content is out there but way out of the reach of the intended audience. For content to be readily available, it has to be properly organized into different categories and structured into an appropriate format using metadata. This paper addresses this issue by proposing an automated approach using ontology-based document classification. The paper presents a case study and describes how our proposed ontology model can be used to classify educational documents into predefined categories.

  • 3.
    Imran, Ali Shariq
    et al.
    Norwegian University of Science and Technology, Norway.
    Kastrati, Zenun
    Linnaeus University, Faculty of Technology, Department of computer science and media technology (CM).
    Svendsen, Torbjørn Karl
    Norwegian University of Science and Technology (NTNU), Norway.
    Kurti, Arianit
    Linnaeus University, Faculty of Technology, Department of computer science and media technology (CM).
    Text-Independent Speaker ID for Automatic Video Lecture Classification Using Deep Learning2019In: Proceedings of the 2019 5th International Conference on Computing and Artificial Intelligence, April 19-22, 2019, Bali, Indonesia, ACM Publications, 2019, p. 175-180Conference paper (Refereed)
    Abstract [en]

    This paper proposes to use acoustic features employing deep neural network (DNN) and convolutional neural network (CNN) models for classifying video lectures in a massive open online course (MOOC). The models exploit the voice pattern of the lecturer for identification and for classifying the video lecture according to the right speaker category. Filter bank and Mel frequency cepstral coefficient (MFCC) feature along with first and second order derivatives (Δ/ΔΔ) are used as input features to the proposed models. These features are extracted from the speech signal which is obtained from the video lectures by separating the audio from the video using FFmpeg.

    The deep learning models are evaluated using precision, recall, and F1 score and the obtained accuracy is compared for both acoustic features with traditional machine learning classifiers for speaker identification. A significant improvement of 3% to 7% classification accuracy is achieved over the DNN and twice to that of shallow machine learning classifiers for 2D-CNN with MFCC. The proposed 2D-CNN model with an F1 score of 85.71% for text-independent speaker identification makes it plausible to use speaker ID as a classification approach for organizing video lectures automatically in a MOOC setting.

  • 4. Kastrati, Z.
    et al.
    Imran, A. S.
    Document image classification using SEMCON2015In: 2015 20th Symposium on Signal Processing, Images and Computer Vision (STSIVA), 2015, p. 1-6Conference paper (Refereed)
  • 5. Kastrati, Z.
    et al.
    Imran, A. S.
    Yayilgan, S. Y.
    SEMCON: Semantic and contextual objective metric2015In: Proceedings of the 2015 IEEE 9th International Conference on Semantic Computing (IEEE ICSC 2015), IEEE, 2015, p. 65-68Conference paper (Refereed)
  • 6. Kastrati, Zenun
    et al.
    Imran, Ali Sahriq
    Yayilgan, Sule Yildirim
    An Improved Concept Vector Space Model for Ontology Based Classification2015In: 2015 11th International Conference on Signal-Image Technology Internet-Based Systems (SITIS), 2015, p. 240-245Conference paper (Refereed)
  • 7.
    Kastrati, Zenun
    et al.
    Norwegian University of Science and Technology, Norway.
    Imran, Ali Shariq
    Norwegian University of Science and Technology, Norway.
    Performance analysis of machine learning classifiers on improved concept vector space models2019In: Future generations computer systems, ISSN 0167-739X, E-ISSN 1872-7115, Vol. 96, p. 552-562Article in journal (Refereed)
    Abstract [en]

    This paper provides a comprehensive performance analysis of parametric and non-parametric machine learning classifiers including a deep feed-forward multi-layer perceptron (MLP) network on two variants of improved Concept Vector Space (iCVS) model. In the first variant, a weighting scheme enhanced with the notion of concept importance is used to assess weight of ontology concepts. Concept importance shows how important a concept is in an ontology and it is automatically computed by converting the ontology into a graph and then applying one of the Markov based algorithms. In the second variant of iCVS, concepts provided by the ontology and their semantically related terms are used to construct concept vectors in order to represent the document into a semantic vector space. We conducted various experiments using a variety of machine learning classifiers for three different models of document representation. The first model is a baseline concept vector space (CVS) model that relies on an exact/partial match technique to represent a document into a vector space. The second and third model is an iCVS model that employs an enhanced concept weighting scheme for assessing weights of concepts (variant 1), and the acquisition of terms that are semantically related to concepts of the ontology for semantic document representation (variant 2), respectively. Additionally, a comparison between seven different classifiers is performed for all three models using precision, recall, and F1 score. Results for multiple configurations of deep learning architecture are obtained by varying the number of hidden layers and nodes in each layer, and are compared to those obtained with conventional classifiers. The obtained results show that the classification performance is highly dependent upon the choice of a classifier, and that the Random Forest, Gradient Boosting, and Multilayer Perceptron are among the classifiers that performed rather well for all three models.

  • 8.
    Kastrati, Zenun
    et al.
    Linnaeus University, Faculty of Technology, Department of computer science and media technology (CM).
    Imran, Ali Shariq
    Norwegian University of Science and Technology, Norway.
    Kurti, Arianit
    Integrating word embeddings and document topics with deep learning in a video classification framework2019In: Pattern Recognition Letters, ISSN 0167-8655, E-ISSN 1872-7344, Vol. 128, p. 85-92Article in journal (Refereed)
    Abstract [en]

    The advent of MOOC platforms brought an abundance of video educational content that made the selection of best fitting content for a specific topic a lengthy process. To tackle this challenge in this paper we report our research efforts of using deep learning techniques for managing and classifying educational content for various search and retrieval applications in order to provide a more personalized learning experience. In this regard, we propose a framework which takes advantages of feature representations and deep learning for classifying video lectures in a MOOC setting to aid effective search and retrieval. The framework consists of three main modules. The first module called pre-processing concerns with video-to-text conversion. The second module is transcript representation which represents text in lecture transcripts into vector space by exploiting different representation techniques including bag-of-words, embeddings, transfer learning, and topic modeling. The final module covers classifiers whose aim is to label video lectures into the appropriate categories. Two deep learning models, namely feed-forward deep neural network (DNN) and convolutional neural network (CNN) are examined as part of the classifier module. Multiple simulations are carried out on a large-scale real dataset using various feature representations and classification techniques to test and validate the proposed framework.

  • 9.
    Kastrati, Zenun
    et al.
    Linnaeus University, Faculty of Technology, Department of computer science and media technology (CM).
    Imran, Ali Shariq
    Norwegian University of Science and Technology - NTNU, Norway.
    Kurti, Arianit
    Linnaeus University, Faculty of Technology, Department of computer science and media technology (CM).
    Transfer Learning to Timed Text Based Video Classification Using CNN2019In: Proceedings of the 9th International Conference on Web Intelligence, Mining and Semantics, Seoul, South Korea: ACM Publications, 2019, article id 23Conference paper (Refereed)
    Abstract [en]

    Open educational video resources are gaining popularity with a growing number of massive open online courses (MOOCs). This has created a niche for content providers to adopt effective solutions in automatically organizing and structuring of educational resources for maximum visibility. Recent advances in deep learning techniques are proving useful in managing and classifying resources into appropriate categories. This paper proposes one such convolutional neural network (CNN) model for classifying video lectures in a MOOC setting using a transfer learning approach. The model uses a time-aligned text transcripts corresponding to video lectures from six broader subject categories. Video lectures and their corresponding transcript dataset is gathered from the Coursera MOOC platform. Two different CNN models are proposed: i) CNN based classification using embeddings learned from our MOOC dataset, ii) CNN based classification using transfer learning. Word embeddings generated from two well known state-of-the-art pre-trained models Word2Vec and GloVe, are used in the transfer learning approach for the second case.

    The proposed CNN models are evaluated using precision, recall, and F1 score and the obtained performance is compared with both conventional and deep learning classifiers. The proposed CNN models have an F1 score improvement of 10-22 percentage points over DNN and conventional classifiers

  • 10. Kastrati, Zenun
    et al.
    Imran, Ali Shariq
    Yayilgan, Sule Yildirim
    A General Framework for Text Document Classification Using SEMCON and ACVSR2015In: Human Interface and the Management of Information. Information and Knowledge Design, Springer International Publishing , 2015, p. 310-319Conference paper (Refereed)
    Abstract [en]

    The text document classification employs either text based approach or semantic based approach to index and retrieve text documents. The former uses keywords and therefore provides limited capabilities to capture and exploit the conceptualization involved in user information needs and content meanings. The latter aims to solve these limitations using content meanings, rather than keywords. More formally, the semantic based approach uses the domain ontology to exploit the content meanings of a particular domain. This approach however has some drawbacks. It lacks enrichment of ontology concepts with new lexical resources and evaluation of the importance indicated by weights of those concepts. Therefore to address these issues, this paper proposes a new ontology based text document classification framework. The proposed framework incorporates a newly developed objective metric calledSEMCON to enrich the domain ontology with new concepts by combining contextual as well as semantic information of a term within a text document. The framework also introduces a new approach to automatically estimate the importance of ontology concepts which is indicated by the weights of these concepts, and to enhance the concept vector space model using automatically estimated weights.

  • 11. Kastrati, Zenun
    et al.
    Imran, Ali Shariq
    Yayilgan, Sule Yildirim
    Building Domain Ontologies for Hyperlinked Multimedia Pedagogical Platforms2014In: HCI International 2014 - Posters’ Extended Abstracts, Springer International Publishing , 2014, p. 95-100Conference paper (Refereed)
    Abstract [en]

    This paper examines building of the course ontology for describing and organizing hyperlinked pedagogical content. The ontology is used to structure and classify multimedia learning objects (MLO) in hyperlinked pedagogical platform called HIP, and to assist students to search for lectures and other teaching materials in a reasonable time and more efficiently. In addition, this paper proposes a new approach to improve the classification performance by enhancing the information representation model using concepts from the pedagogical course domain ontology. The model will automatically estimate weight of concepts within the ontology, and it will combine the weight with concepts’ importance which is calculated using Term Frequency Inverse Document Frequency – tf*idf algorithm. This paper is a work in progress. We are in process of creating and implementing the course ontology and an experiment will be conducted to evaluate the classification performance in terms of efficiency and effectiveness for the approach proposed in this paper.

  • 12.
    Kastrati, Zenun
    et al.
    Norwegian University of Science and Technology, Norway.
    Imran, Ali Shariq
    Norwegian University of Science and Technology, Norway.
    Yayilgan, Sule Yildirim
    Norwegian University of Science and Technology, Norway.
    The impact of deep learning on document classification using semantically rich representations2019In: Information Processing & Management, ISSN 0306-4573, E-ISSN 1873-5371, Vol. 56, no 5, p. 1618-1632Article in journal (Refereed)
    Abstract [en]

    This paper presents a semantically rich document representation model for automatically classifying financial documents into predefined categories utilizing deep learning. The model architecture consists of two main modules including document representation and document classification. In the first module, a document is enriched with semantics using background knowledge provided by an ontology and through the acquisition of its relevant terminology. Acquisition of terminology integrated to the ontology extends the capabilities of semantically rich document representations with an in depth-coverage of concepts, thereby capturing the whole conceptualization involved in documents. Semantically rich representations obtained from the first module will serve as input to the document classification module which aims at finding the most appropriate category for that document through deep learning. Three different deep learning networks each belonging to a different category of machine learning techniques for ontological document classification using a real-life ontology are used. Multiple simulations are carried out with various deep neural networks configurations, and our findings reveal that a three hidden layer feedforward network with 1024 neurons obtain the highest document classification performance on the INFUSE dataset. The performance in terms of F1 score is further increased by almost five percentage points to 78.10% for the same network configuration when the relevant terminology integrated to the ontology is applied to enrich document representation. Furthermore, we conducted a comparative performance evaluation using various state-of-the-art document representation approaches and classification techniques including shallow and conventional machine learning classifiers.

  • 13.
    Kastrati, Zenun
    et al.
    Norwegian University of Science and Technology, Norway.
    Imran, Ali Shariq
    Norwegian university of science and technology, Norway.
    Yildirim-Yayilgan, Sule
    Norwegian university of science and technology, Norway.
    A Hybrid Concept Learning Approach to Ontology Enrichment2018In: Innovations, Developments, and Applications of Semantic Web and Information Systems / [ed] Miltiadis D. Lytras, Naif Aljohani, Ernesto Damiani & Kwok Tai Chui, IGI Global, 2018, p. 85-119Chapter in book (Other academic)
  • 14. Kastrati, Zenun
    et al.
    Imran, Ali Shariq
    Yildirim-Yayilgan, Sule
    SEMCON: A Semantic and Contextual Objective Metric for Enriching Domain Ontology Concepts2016In: International Journal on Semantic Web and Information Systems (IJSWIS), Vol. 12, no 2, p. 1-24Article in journal (Refereed)
    Abstract [en]

    This paper presents a novel concept enrichment objective metric combining contextual and semantic information of terms extracted from the domain documents. The proposed metric is called SEMCON which stands for semantic and contextual objective metric. It employs a hybrid learning approach utilizing functionalities from statistical and linguistic ontology learning techniques. The metric also introduced for the first time two statistical features that have shown to improve the overall score ranking of highly relevant terms for concept enrichment. Subjective and objective experiments are conducted in various domains. Experimental results (F1) from computer domain show that SEMCON achieved better performance in contrast to tf*idf, and LSA methods, with 12.2%, 21.8%, and 24.5% improvement over them respectively. Additionally, an investigation into how much each of contextual and semantic components contributes to the overall task of concept enrichment is conducted and the obtained results suggest that a balanced weight gives the best performance.

  • 15. Kastrati, Zenun
    et al.
    Yayilgan, Sule Yildirim
    Improving Document Classification Effectiveness Using Knowledge Exploited by Ontologies2017In: Natural Language Processing and Information Systems, Springer International Publishing , 2017, p. 435-438Conference paper (Refereed)
    Abstract [en]

    In this paper, we propose a new document classification model which utilizes background knowledge gathered by ontologies for document representation. A document is represented using a set of ontology concepts that are acquired by exact matching technique and through identification and extraction of new terms which can be semantically related to these concepts. In addition, a new concept weighting scheme composed of concept relevance and importance is employed by the model to compute weight of concepts. We conducted experiments to test the model and the obtained results showed that a considerable improvement of classification performance is achieved by using our proposed model.

  • 16.
    Kastrati, Zenun
    et al.
    Norwegian University of Science and Technology, Norway.
    Yayilgan, Sule Yildirim
    Norwegian University of Science and Technology, Norway.
    Supervised Ontology-Based Document Classification Model2017In: Proceedings of the International Conference on Compute and Data Analysis, ICCDA'17, ACM Publications, 2017, p. 245-251Conference paper (Refereed)
    Abstract [en]

    Ontology-based document classification relies on background knowledge exploited by ontologies to represent documents. Background knowledge is embedded in a document using the exact matching technique. The basic idea of this technique is to map a term to a concept by searching only the concept labels that explicitly occur in a document. Searching only the presence of concept labels limits the capabilities to capture and exploit the whole conceptualization involved in user information and content meanings. Therefore, to address this limitation, we propose a new document classification model based on ontologies. The proposed model uses background knowledge derived by ontologies for document representation. It associates a document with a set of concepts by not only using the exact matching technique but also by identifying and extracting new terms which can be semantically related to the concepts of ontologies. Additionally, the proposed model employs a new concept weighting technique which computes the weight of a concept using the relevance and the importance of the concept. We conducted several experiments using a real ontology and a dataset to test our proposed model. The results obtained by experiments run on 3 different classification algorithms using the baseline ontology, the improved concept vector space model by using the new concept weighting technique, and the enriched ontology, show that our proposed model achieved a considerable improvement of classification performance.

  • 17. Kastrati, Zenun
    et al.
    Yayilgan, Sule Yildirim
    Hjelsvold, Rune
    Automatically Enriching Domain Ontologies for Document Classification2016In: Proceedings of the 6th International Conference on Web Intelligence, Mining and Semantics, ACM , 2016, p. 29-1Conference paper (Refereed)
  • 18. Kastrati, Zenun
    et al.
    Yayilgan, Sule Yildirim
    Imran, Ali Shariq
    Using Context-Aware and Semantic Similarity Based Model to Enrich Ontology Concepts2015In: Natural Language Processing and Information Systems, Springer International Publishing , 2015, p. 137-143Conference paper (Refereed)
    Abstract [en]

    Domain ontologies are a good starting point to model in a formal way the basic vocabulary of a given domain. However, in order for an ontology to be usable in real applications, it has to be supplemented with lexical resources of this particular domain. The learning process of enriching domain ontologies with new lexical resources employed in the existing approaches takes into account only the contextual aspects of terms and does not consider their semantics. Therefore, this paper proposes a new objective metric namely SEMCON which combines contextual as well as semantic information of terms to enriching the domain ontology with new concepts. The SEMCON defines the context by first computing an observation matrix which exploits the statistical features such as frequency of the occurrence of a term, term’s font type and font size. The semantics is then incorporated by computing a semantic similarity score using lexical database WordNet. Subjective and objective experiments are conducted and results show an improved performance of SEMCON compared with tf*idf and $$\chi ^2$$.

  • 19. Picovici, Dorel
    et al.
    Denieffe, David
    Kastrati, Zenun
    Subjective-based Quality Assessment for Online Games2010In: Proceedings of the 3rd International ICST Conference on Simulation Tools and Techniques, ICST (Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering) , 2010, p. 10-1Conference paper (Refereed)
1 - 19 of 19
CiteExportLink to result list
Permanent link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf