lnu.sePublications
Change search
Link to record
Permanent link

Direct link
BETA
Publications (10 of 49) Show all publications
Kastrati, Z., Imran, A. S. & Kurti, A. (2019). Integrating word embeddings and document topics with deep learning in a video classification framework. Pattern Recognition Letters, 128, 85-92
Open this publication in new window or tab >>Integrating word embeddings and document topics with deep learning in a video classification framework
2019 (English)In: Pattern Recognition Letters, ISSN 0167-8655, E-ISSN 1872-7344, Vol. 128, p. 85-92Article in journal (Refereed) Published
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.

Place, publisher, year, edition, pages
Elsevier, 2019
Keywords
Deep learning, Video classification, Embedding, Document topics, CNN, DNN
National Category
Computer Sciences
Research subject
Computer and Information Sciences Computer Science, Information Systems
Identifiers
urn:nbn:se:lnu:diva-88861 (URN)10.1016/j.patrec.2019.08.019 (DOI)
Available from: 2019-08-30 Created: 2019-08-30 Last updated: 2019-09-06Bibliographically approved
Alsouda, Y., Pllana, S. & Kurti, A. (2019). IoT-based Urban Noise Identification Using Machine Learning: Performance of SVM, KNN, Bagging, and Random Forest. In: Proceedings of the International Conference on Omni-Layer Intelligent Systems (COINS '19): . Paper presented at International Conference on Omni-Layer Intelligent Systems (COINS '19), Crete, Greece — May 05 - 07, 2019 (pp. 62-67). New York: ACM Publications
Open this publication in new window or tab >>IoT-based Urban Noise Identification Using Machine Learning: Performance of SVM, KNN, Bagging, and Random Forest
2019 (English)In: Proceedings of the International Conference on Omni-Layer Intelligent Systems (COINS '19), New York: ACM Publications, 2019, p. 62-67Conference paper, Published paper (Refereed)
Abstract [en]

Noise is any undesired environmental sound. A sound at the same dB level may be perceived as annoying noise or as pleasant music. Therefore, it is necessary to go beyond the state-of-the-art approaches that measure only the dB level and also identify the type of noise. In this paper, we present a machine learning based method for urban noise identification using an inexpensive IoT unit. We use Mel-frequency cepstral coefficients for audio feature extraction and supervised classification algorithms (that is, support vector machine, k-nearest neighbors, bootstrap aggregation, and random forest) for noise classification. We evaluate our approach experimentally with a data-set of about 3000 sound samples grouped in eight sound classes (such as car horn, jackhammer, or street music). We explore the parameter space of the four algorithms to estimate the optimal parameter values for classification of sound samples in the data-set under study. We achieve a noise classification accuracy in the range 88% - 94%.

Place, publisher, year, edition, pages
New York: ACM Publications, 2019
Keywords
bootstrap aggregation (Bagging), internet of things (IoT), k-nearest neighbors (KNN), mel-frequency cepstral coefficients (MFCC), random forest, smart cities, support vector machine (SVM), urban noise
National Category
Computer Systems
Research subject
Computer Science, Software Technology
Identifiers
urn:nbn:se:lnu:diva-81767 (URN)10.1145/3312614.3312631 (DOI)2-s2.0-85066804134 (Scopus ID)978-1-4503-6640-3 (ISBN)
Conference
International Conference on Omni-Layer Intelligent Systems (COINS '19), Crete, Greece — May 05 - 07, 2019
Funder
Knowledge Foundation, 20150088, 20150259
Available from: 2019-04-09 Created: 2019-04-09 Last updated: 2019-08-29Bibliographically approved
Memeti, S., Pllana, S., Ferati, M., Kurti, A. & Jusufi, I. (2019). IoTutor: How Cognitive Computing Can Be Applied to Internet of Things Education. In: Leon Strous and Vinton G. Cerf (Ed.), : . Paper presented at IFIPIoT 2018 (pp. 1-16). Springer
Open this publication in new window or tab >>IoTutor: How Cognitive Computing Can Be Applied to Internet of Things Education
Show others...
2019 (English)Conference paper, Published paper (Refereed)
Abstract [en]

We present IoTutor that is a cognitive computing solution for education of students in the IoT domain. We implement the IoTutor as a platform-independent web-based application that is able to interact with users via text or speech using natural language. We train the IoTutor with selected scientific publications relevant to the IoT education. To investigate users' experience with the IoTutor, we ask a group of students taking an IoT master level course at the Linnaeus University to use the IoTutor for a period of two weeks. We ask students to express their opinions with respect to the attractiveness, perspicuity, efficiency, stimulation, and novelty of the IoTutor. The evaluation results show a trend that students express an overall positive attitude towards the IoTutor with majority of the aspects rated higher than the neutral value.

Place, publisher, year, edition, pages
Springer, 2019
Series
IFIP Advances in Information and Communication Technology, ISSN 1868-4238 ; 548
Keywords
Internet of Things (IoT), education, cognitive computing, IBM Watson
National Category
Computer Systems
Research subject
Computer and Information Sciences Computer Science
Identifiers
urn:nbn:se:lnu:diva-80835 (URN)10.1007/978-3-030-15651-0_18 (DOI)2-s2.0-85064686693 (Scopus ID)978-3-030-15651-0 (ISBN)978-3-030-15650-3 (ISBN)
Conference
IFIPIoT 2018
Funder
Knowledge Foundation, 20150088, 20150259
Available from: 2019-02-26 Created: 2019-02-26 Last updated: 2019-08-29Bibliographically approved
Imran, A. S., Kastrati, Z., Svendsen, T. K. & Kurti, A. (2019). Text-Independent Speaker ID for Automatic Video Lecture Classification Using Deep Learning. In: Proceedings of the 2019 5th International Conference on Computing and Artificial Intelligence, April 19-22, 2019, Bali, Indonesia: . Paper presented at 5th International Conference on Computing and Artificial Intelligence, April 19-22, 2019, Bali, Indonesia (pp. 175-180). ACM Publications
Open this publication in new window or tab >>Text-Independent Speaker ID for Automatic Video Lecture Classification Using Deep Learning
2019 (English)In: 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, Published 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.

Place, publisher, year, edition, pages
ACM Publications, 2019
Keywords
2D-CNN, DNN, MFCC Filter banks, MOOC, Speaker identification, deep learning, video classification
National Category
Computer Sciences
Research subject
Computer and Information Sciences Computer Science, Computer Science
Identifiers
urn:nbn:se:lnu:diva-88114 (URN)10.1145/3330482.3330508 (DOI)978-1-4503-6106-4 (ISBN)
Conference
5th International Conference on Computing and Artificial Intelligence, April 19-22, 2019, Bali, Indonesia
Available from: 2019-08-19 Created: 2019-08-19 Last updated: 2019-09-04Bibliographically approved
Kastrati, Z., Imran, A. S. & Kurti, A. (2019). Transfer Learning to Timed Text Based Video Classification Using CNN. In: Proceedings of the 9th International Conference on Web Intelligence, Mining and Semantics: . Paper presented at 9th International Conference on Web Intelligence, Mining and Semantics, 26-28 June, Republic of Korea. Seoul, South Korea: ACM Publications, Article ID 23.
Open this publication in new window or tab >>Transfer Learning to Timed Text Based Video Classification Using CNN
2019 (English)In: Proceedings of the 9th International Conference on Web Intelligence, Mining and Semantics, Seoul, South Korea: ACM Publications, 2019, article id 23Conference paper, Published 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

Place, publisher, year, edition, pages
Seoul, South Korea: ACM Publications, 2019
National Category
Information Systems, Social aspects
Research subject
Computer and Information Sciences Computer Science, Computer Science
Identifiers
urn:nbn:se:lnu:diva-88590 (URN)10.1145/3326467.3326483 (DOI)978-1-4503-6190-3 (ISBN)
Conference
9th International Conference on Web Intelligence, Mining and Semantics, 26-28 June, Republic of Korea
Available from: 2019-08-24 Created: 2019-08-24 Last updated: 2019-09-04Bibliographically approved
Alsouda, Y., Pllana, S. & Kurti, A. (2018). A Machine Learning Driven IoT Solution for Noise Classification in Smart Cities. In: Machine Learning Driven Technologies and Architectures for Intelligent Internet of Things (ML-IoT), August 28, 2018, Prague, Czech Republic: . Paper presented at Machine Learning Driven Technologies and Architectures for Intelligent Internet of Things (ML-IoT), August 28, 2018, Prague, Czech Republic (pp. 1-6). Euromicro
Open this publication in new window or tab >>A Machine Learning Driven IoT Solution for Noise Classification in Smart Cities
2018 (English)In: Machine Learning Driven Technologies and Architectures for Intelligent Internet of Things (ML-IoT), August 28, 2018, Prague, Czech Republic, Euromicro , 2018, p. 1-6Conference paper, Published paper (Refereed)
Abstract [en]

We present a machine learning based method for noise classification using a low-power and inexpensive IoT unit. We use Mel-frequency cepstral coefficients for audio feature extraction and supervised classification algorithms (that is, support vector machine and k-nearest neighbors) for noise classification. We evaluate our approach experimentally with a dataset of about 3000 sound samples grouped in eight sound classes (such as, car horn, jackhammer, or street music). We explore the parameter space of support vector machine and k-nearest neighbors algorithms to estimate the optimal parameter values for classification of sound samples in the dataset under study. We achieve a noise classification accuracy in the range 85% -- 100%. Training and testing of our k-nearest neighbors (k = 1) implementation on Raspberry Pi Zero W is less than a second for a dataset with features of more than 3000 sound samples.

Place, publisher, year, edition, pages
Euromicro, 2018
Keywords
urban noise, smart cities, support vector machine (SVM), k-nearest neighbors (KNN), mel-frequency cepstral coefficients (MFCC), internet of things (IoT)
National Category
Computer Systems
Research subject
Computer and Information Sciences Computer Science, Computer Science
Identifiers
urn:nbn:se:lnu:diva-81672 (URN)
Conference
Machine Learning Driven Technologies and Architectures for Intelligent Internet of Things (ML-IoT), August 28, 2018, Prague, Czech Republic
Available from: 2019-04-04 Created: 2019-04-04 Last updated: 2019-05-20Bibliographically approved
Dalipi, F., Ferati, M. & Kurti, A. (2018). Integrating MOOCs in Regular Higher Education: Challenges and Opportunities from a Scandinavian Perspective. In: Panayiotis Zaphiris and Andri Ioannou (Ed.), Learning and Collaboration Technologies: Design, Development and Technological Innovation. LCT 2018. Paper presented at 5th International Conference on Learning and Collaboration Technologies, LCT 2018 Held as Part of HCI International 2018; Las Vegas; United States; 15-20 July 2018 (pp. 193-204). Springer, 10924
Open this publication in new window or tab >>Integrating MOOCs in Regular Higher Education: Challenges and Opportunities from a Scandinavian Perspective
2018 (English)In: Learning and Collaboration Technologies: Design, Development and Technological Innovation. LCT 2018 / [ed] Panayiotis Zaphiris and Andri Ioannou, Springer, 2018, Vol. 10924, p. 193-204Conference paper, Published paper (Refereed)
Abstract [en]

MOOCs are increasingly being considered by universities as an integral part of their curriculum. Nevertheless, there are several challenges that to some extent slow this process, where the most important one is the accreditation challenges and financing. These challenges are particularly important in the context of universities in Scandinavian countries where education is mostly free. In order to gain more insights on the status of proliferation of MOOCs in Scandinavian universities and understand any specific challenges, we conducted a study by analyzing two sources of data: research publications and university websites. Further on, these data have been analyzed using a framework that differentiates and categorizes MOOCs in terms of accreditation and scalability. As a result of this analysis, we have identified the remaining challenges as well as a number of opportunities regarding the full integration of MOOCs in the educational system of the Scandinavian Higher Education Institutions.

Place, publisher, year, edition, pages
Springer, 2018
Series
Lecture Notes in Computer Science, ISSN 0302-9743 ; 10924
Keywords
Denmark, Higher education, MOOCs, Norway, Online learning, Opportunities, Scandinavia, Sweden, Artificial intelligence, Computer science, Computers, Accreditation
National Category
Information Systems, Social aspects
Research subject
Computer and Information Sciences Computer Science, Information Systems
Identifiers
urn:nbn:se:lnu:diva-77804 (URN)10.1007/978-3-319-91743-6_15 (DOI)2-s2.0-85050572535 (Scopus ID)9783319917429 (ISBN)
Conference
5th International Conference on Learning and Collaboration Technologies, LCT 2018 Held as Part of HCI International 2018; Las Vegas; United States; 15-20 July 2018
Funder
Knowledge Foundation, 67110033
Available from: 2018-09-14 Created: 2018-09-14 Last updated: 2018-10-04Bibliographically approved
Ahmedi, F., Ahmedi, L., O'Flynn, B., Kurti, A., Tahirsylaj, S., Bytyçi, E., . . . Salihu, A. (2018). InWaterSense: An Intelligent Wireless Sensor Network for Monitoring Surface Water Quality to a River in Kosovo. Paper presented at Hershey, PA, USA. International Journal of Agricultural and Environmental Information Systems, 9(1), 39-61
Open this publication in new window or tab >>InWaterSense: An Intelligent Wireless Sensor Network for Monitoring Surface Water Quality to a River in Kosovo
Show others...
2018 (English)In: International Journal of Agricultural and Environmental Information Systems, ISSN 1947-3192, Vol. 9, no 1, p. 39-61Article in journal (Refereed) Published
Abstract [en]

A shift in water monitoring approach from traditional grab sampling to novel wireless sensors is gaining in popularity not only among researchers but also in the market. These latest technologies readily enable numerous advantageous monitoring arrangements like remote, continuous, real-time, and spatially-dense and broad in coverage measurements, and identification of long-term trends of parameters of interest. Thus, a WSN system is implemented in a river in Kosovo as part of the InWaterSense project to monitor its water quality parameters. It is one of the first state of the art technology demonstration systems of its kind in the domain of water monitoring in developing countries like Kosovo. Water quality datasets are transmitted at pre-programmed intervals from sensing stations deployed in the river to the server at university via the GPRS network. Data is then made available through a portal to different target groups (policy-makers, water experts, and citizens). Moreover, the InWaterSense system behaves intelligently like staying in line with water quality regulatory standards.

Place, publisher, year, edition, pages
IGI Global, 2018
National Category
Other Engineering and Technologies not elsewhere specified
Research subject
Computer and Information Sciences Computer Science, Computer Science
Identifiers
urn:nbn:se:lnu:diva-68718 (URN)10.4018/IJAEIS.2018010103 (DOI)000429506200003 ()2-s2.0-85034016443 (Scopus ID)
Conference
Hershey, PA, USA
Projects
InWaterSense
Available from: 2017-11-11 Created: 2017-11-11 Last updated: 2019-08-29Bibliographically approved
Chow, J. A., Törnros, M. E., Waltersson, M., Richard, H., Kusoffsky, M., Lundström, C. F. & Kurti, A. (2017). A design study investigating augmented reality and photograph annotation in a digitalized grossing workstation. Journal of Pathology Informatics, 8, Article ID 31.
Open this publication in new window or tab >>A design study investigating augmented reality and photograph annotation in a digitalized grossing workstation
Show others...
2017 (English)In: Journal of Pathology Informatics, ISSN 2229-5089, E-ISSN 2153-3539, Vol. 8, article id 31Article in journal (Refereed) Published
Abstract [en]

Context: Within digital pathology, digitalization of the grossing procedure has been relatively underexplored in comparison to digitalization of pathology slides. Aims: Our investigation focuses on the interaction design of an augmented reality gross pathology workstation and refining the interface so that information and visualizations are easily recorded and displayed in a thoughtful view. Settings and Design: The work in this project occurred in two phases: the first phase focused on implementation of an augmented reality grossing workstation prototype while the second phase focused on the implementation of an incremental prototype in parallel with a deeper design study. Subjects and Methods: Our research institute focused on an experimental and “designerly” approach to create a digital gross pathology prototype as opposed to focusing on developing a system for immediate clinical deployment. Statistical Analysis Used: Evaluation has not been limited to user tests and interviews, but rather key insights were uncovered through design methods such as “rapid ethnography” and “conversation with materials”. Results: We developed an augmented reality enhanced digital grossing station prototype to assist pathology technicians in capturing data during examination. The prototype uses a magnetically tracked scalpel to annotate planned cuts and dimensions onto photographs taken of the work surface. This article focuses on the use of qualitative design methods to evaluate and refine the prototype. Our aims were to build on the strengths of the prototype's technology, improve the ergonomics of the digital/physical workstation by considering numerous alternative design directions, and to consider the effects of digitalization on personnel and the pathology diagnostics information flow from a wider perspective. A proposed interface design allows the pathology technician to place images in relation to its orientation, annotate directly on the image, and create linked information. Conclusions: The augmented reality magnetically tracked scalpel reduces tool switching though limitations in today's augmented reality technology fall short of creating an ideal immersive workflow by requiring the use of a monitor. While this technology catches up, we recommend focusing efforts on enabling the easy creation of layered, complex reports, linking, and viewing information across systems. Reflecting upon our results, we argue for digitalization to focus not only on how to record increasing amounts of data but also how these data can be accessed in a more thoughtful way that draws upon the expertise and creativity of pathology professionals using the systems.

Place, publisher, year, edition, pages
Medknow Publications, 2017
Keywords
Augmented reality, design methods, gross pathology, human–computer interaction, interface design, visualization
National Category
Biomedical Laboratory Science/Technology
Research subject
Computer and Information Sciences Computer Science; Computer Science, Information and software visualization
Identifiers
urn:nbn:se:lnu:diva-68016 (URN)10.4103/jpi.jpi_13_17 (DOI)
Funder
VINNOVA
Available from: 2017-09-18 Created: 2017-09-18 Last updated: 2019-05-20Bibliographically approved
Kurti, A. & Dalipi, F. (2017). Bridging the Gap between Academia and Industry: Lessons Learned from a Graduate IT Professional Development Program. In: Gregory T. Papanikos (Ed.), Abstract Book: 2nd Annual International Conference on Engineering Education & Teaching, 5-8 June 2017, Athens, Greece. Paper presented at 2nd Annual International Conference on Engineering Education & Teaching, 5-8 June 2017, Athens (pp. 27-27). Athens
Open this publication in new window or tab >>Bridging the Gap between Academia and Industry: Lessons Learned from a Graduate IT Professional Development Program
2017 (English)In: Abstract Book: 2nd Annual International Conference on Engineering Education & Teaching, 5-8 June 2017, Athens, Greece / [ed] Gregory T. Papanikos, Athens, 2017, p. 27-27Conference paper, Oral presentation with published abstract (Other academic)
Abstract [en]

The rapid advances of technologies, constantly brings new demands for new skills and expertise of the professionals in IT industry. There is a constant need for people that have in-depth understanding and know how to develop the new innovative services using these new technologies. In these settings, the real challenge is how to find the right persons with the right education in an industry where the in-thing yesterday may be out-of-date tomorrow? To add to this challenge, universities are still “increasingly stove-piped in highly specialized disciplinary fields” (Hurlburt et al., 2010) as well as there is a lack of flexibility for the professionals to have their competences developed. All this points out the great challenges that universities are facing for alignment between academic development within degree curricula and the requirements that industry demands for their specific needs (Falcone et al. 2014). In this research effort we report our experiences from an ongoing Graduate Professional Development Program where we address these challenges through a co-creation process with IT industry based on open innovation. Through this model we bring together research expertise, academic experience and experts from industry in a collaborative process for developing courses to suit the current needs of IT professionals. As an outcome of this process, the course content is tailor-made, as well as everything else in connection, such as: bite-size modules, adjustable pace, open and online educational resources, as well as a flipped classroom approach to teaching. As a result, we have developed and provided so far five courses that have been very well accepted by the IT professional. Thus, in this paper we aim to provide some insights on approaches for facilitating continuous competence development plans for IT professionals within regular university educational offer. 

Place, publisher, year, edition, pages
Athens: , 2017
National Category
Computer and Information Sciences
Research subject
Computer and Information Sciences Computer Science, Computer Science
Identifiers
urn:nbn:se:lnu:diva-67486 (URN)978-960-598-128-0 (ISBN)
Conference
2nd Annual International Conference on Engineering Education & Teaching, 5-8 June 2017, Athens
Funder
Knowledge Foundation
Available from: 2017-08-29 Created: 2017-08-29 Last updated: 2019-05-20Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0003-0512-6350

Search in DiVA

Show all publications