Papers by CHRISTOS SKOURLAS
Evaluating Citizen Comments in Public Consultations Using Data Mining
In the present paper, in the context of a pilot application, comments that appear on the Consulta... more In the present paper, in the context of a pilot application, comments that appear on the ConsultationWebsite of the Ministry of Development and Investments of Greece, concerning the draft provisions of the Greek General Commercial Registry, were collected and studied. Algorithms that use various data mining methods were applied in order to analyze citizens’ comments. The data set was processed with the algorithms Apriori, Naive Bayes, J4.8 and K-Means through Weka software. The aim of this study is to classify citizens’ comments, discover behavioral patterns with the use of association rules, identify groups with similar comments and recommendations using the clustering method, and study whether changes to the law are proposed. The study of the experimental results showed important findings about how citizens react towards law changes and how they can assist the public administration change old timeconsuming practices, adopt new policies, and improve the public administrations’ decision making.
A Framework for Communities of Practice in Radiotherapy
Springer proceedings in business and economics, Sep 27, 2016
This paper focuses on the Communities of Practice (CoPs) which is seen as a specific knowledge ma... more This paper focuses on the Communities of Practice (CoPs) which is seen as a specific knowledge management strategy. We examine how and why CoPs are established in radiotherapy. Based on experience in the radiotherapy practice we discuss a conceptual framework for the assessment of CoPs in radiotherapy.
The Maximum entropy (ME) approach has been extensively used for various natural language processi... more The Maximum entropy (ME) approach has been extensively used for various natural language processing tasks, such as language modeling, part-of-speech tagging, text segmentation and text classification. Previous work in text classification has been done using maximum entropy modeling with binary-valued features or counts of feature words. In this work, we present a method to apply Maximum Entropy modeling for text classification in a different way it has been used so far, using weights for both to select the features of the model and to emphasize the importance of each one of them in the classification task. Using the X square test to assess the contribution of each candidate feature from the obtained X square values we rank the features and the most prevalent of them, those which are ranked with the higher X square scores, they are used as the selected features of the model. Instead of using Maximum Entropy modeling in the classical way, we use the X square values to weight the features of the model and give thus a different importance to each one of them. The method has been evaluated on Reuters-21578 dataset for test classification tasks, giving very promising results and performing comparable to some of the "state of the art" systems in the classification field.
An integrated system for predicting students’ academic performance in smart universities
Due to the current situation with Coronavirus (COVID-19) the attendance of students in the academ... more Due to the current situation with Coronavirus (COVID-19) the attendance of students in the academic life has changed and the educational process has been driven towards smart educational environments. Higher educational Institutes invest significant resources in reforming their educational programs so that it will support distance learning using asynchronous or synchronous methodologies and tools. In this work, we propose the development of a student profile using data from both asynchronous and synchronous e-learning platforms, using a multi-layered neural network in order to classify students’ performance. A neural network is compared against Support Vector Machines, k-Nearest Neighbour and decision trees. The results indicate that the Neural network achieves better accuracy than the others, so using our methodology the instructors or the policy makers of the institute will be able to keep informed about the performance of the students, or take the appropriate actions in order to prevent student failure or low participation.
Academic Data Derived from a University E-government analytic platform: An Educational Data Mining Approach
Data in Brief, Aug 1, 2023
Knowledge Management Techniques in Emotion-Based Music Recommendation Systems
Springer eBooks, 2021
In this paper two statistical methods for extracting collocations from text corpora written in Mo... more In this paper two statistical methods for extracting collocations from text corpora written in Modern Greek are described, the mean and variance method and a method based on the X 2 test. The mean and variance method calculates distances ("offsets") between words in a corpus and looks for specific patterns of distance. The X 2 test is combined with the formulation of a null hypothesis H 0 for a sample of occurrences and we check if there are associations between the words. The X 2 testing does not assume that the words in the corpus have normally distributed probabilities and hence it seems to be more flexible. The two methods extract interesting collocations that are useful in various applications e.g. computational lexicography, language generation and machine translation.
Springer proceedings in business and economics, Sep 27, 2016
This paper investigates a framework for Secure Retrieval and Dissemination of Information (text a... more This paper investigates a framework for Secure Retrieval and Dissemination of Information (text and image) in Distributed and Wireless Environments (SECRET_DIDWE). Our research focuses on the evaluation and integration of Information Retrieval techniques. Text Retrieval and Content Based Image Retrieval techniques are mainly studied for use in medical environment. The framework is based on a wireless architecture to enable authorized medical personnel to access medical records in a secure and transparent manner, utilizing an agent based architecture. A policy-based architecture is part of the framework for utilizing wireless sensor devices, advanced network topologies and software agents to enable remote monitoring of patients and elderly people. These technologies can be used to achieve continuous monitoring of a patient's condition. Medical information classification based on neural network techniques of SVM (Support Vector Machines) type for supporting diagnosis is also incorporated into the framework.
Ranking tokens with class label frequencies for medical article classification
In this paper, a new method for medical article classification is proposed based on exploiting in... more In this paper, a new method for medical article classification is proposed based on exploiting information from local and global class label frequencies in training corpus. The proposed method partially overcomes the low accuracy rate of KNN classifier. First, it uses a lexical approach to identify tokens in the medical document article and then, it uses local and global class label frequencies in a sophisticated way similar to traditional tf-idf weighting scheme to devise the weighted function in classification process. The evaluation experiments on the collection of medical documents, called Ohsumed, show that the method proposed here significantly outperforms traditional KNN classification.
An integrated system for predicting students’ academic performance in smart universities
24th Pan-Hellenic Conference on Informatics
Due to the current situation with Coronavirus (COVID-19) the attendance of students in the academ... more Due to the current situation with Coronavirus (COVID-19) the attendance of students in the academic life has changed and the educational process has been driven towards smart educational environments. Higher educational Institutes invest significant resources in reforming their educational programs so that it will support distance learning using asynchronous or synchronous methodologies and tools. In this work, we propose the development of a student profile using data from both asynchronous and synchronous e-learning platforms, using a multi-layered neural network in order to classify students’ performance. A neural network is compared against Support Vector Machines, k-Nearest Neighbour and decision trees. The results indicate that the Neural network achieves better accuracy than the others, so using our methodology the instructors or the policy makers of the institute will be able to keep informed about the performance of the students, or take the appropriate actions in order to prevent student failure or low participation.
Database SQL transactions and learning by verifying in practice
Proceedings of the 19th Panhellenic Conference on Informatics
In this paper, we present and briefly discuss the case of the DBTechNet, the "DBTech VET Tea... more In this paper, we present and briefly discuss the case of the DBTechNet, the "DBTech VET Teachers" project, and the DBTechNet Course Module on Database SQL Transactions. Data collection and data analysis are mainly based on the pilot offering of the course to an international audience in Greece. Evaluation results, and the types of participants who have attended the course offerings, are presented as target groups for future activities.
IoT-enabled fall verification of elderly and impaired people in smart cities
Proceedings of the 22nd Pan-Hellenic Conference on Informatics
Internet of Things (IoT) technology enables the transformation of urban habitation towards Smart ... more Internet of Things (IoT) technology enables the transformation of urban habitation towards Smart Cities. 70% of human habitation will live in Smart Cities by 2050. IoT is the backbone technology of Smart Cities enabling new services through certain structural parameters. One such parameter is the smart way of living which aims in providing all the necessary technology required to transform human lives to a viable environment of prosperity. Elderly people is a parameter of highly importance in Smart Cities. IoT provides the technical opportunities to elderly and/or impaired people to live and prosper in the context of smart way of living. In this paper, we propose a model for validating elderly and/or impaired people fall. The proposed model is able to define whether the event is a free fall or a bending of the individual towards the floor surface. The IoT equipment we use is the wearable in clothing altimeter sensor.
Towards a framework for increasing Students' participation in Quality Assurance procedures
24th Pan-Hellenic Conference on Informatics
The purpose of this research is to propose and discuss ways of increasing the participation of st... more The purpose of this research is to propose and discuss ways of increasing the participation of students in Quality Assurance (QA) activities at all levels, in their University, through cooperation and mobility programs in Higher Education, etc. First, the literature is reviewed and a comparative analysis of the various methods of increasing student participation in QA activities is conducted. Then we propose the creation and use of an Information System (IS) based on the research results. The IS will help students to understand that not only is their opinion recorded, but also that their views have an impact on the University. Eventually, we discuss the design of the four subsystems of the proposed IS: The “Rate your Lecturer” system, the “Evaluation based on Social Networking” system, the “Students’ view on critical issues” system, and the “Sentiment analysis of students’ views on critical issues” system.
Στην παρούσα διπλωματική εργασία στο πλαίσιο του διεπιστημονικού τομέα της διαχείρισης γνώσης (Kn... more Στην παρούσα διπλωματική εργασία στο πλαίσιο του διεπιστημονικού τομέα της διαχείρισης γνώσης (Knowledge Management) γίνεται μελέτη των συστημάτων μητρώου ειδικών (Yellow Pages Systems). Τα συστήματα αυτά διευκολύνουν την αναζήτηση ειδικών (experts), δηλαδή ανθρώπων εντός ή εκτός οργανισμού με την κατάλληλη εξειδίκευση και τα κατάλληλα προσόντα. Στόχος είναι η επικοινωνία μαζί τους και η συνεργασία. Στο πλαίσιο της εργασίας αναπτύσσεται πιλοτική εφαρμογή ενός τέτοιου συστήματος και η εφαρμογή δοκιμάζεται με δεδομένα ελέγχου (datasets). Γίνεται ανάπτυξη ενός συστήματος για την αποθήκευση και επεξεργασία των δεδομένων. Κατά την επεξεργασία των δεδομένων χρησιμοποιούνται blocking και filtering τεχνικές για να ομαδοποιηθούν τα δεδομένα. Για την υλοποίηση μελετήθηκαν ως κύρια εργαλεία το εργαλείο ΜonetDB και η γλώσσα προγραμματισμού python. Το εργαλείο ΜonetDB είναι σύστημα διαχείρισης βάσης δεδομένων ανοιχτού κώδικα με προσανατολισμό στις στήλες και είναι πολύ αποτελεσματικό στη διαχείρ...
Η ανάπτυξη των Τεχνολογιών Πληροφορικής και Επικοινωνιών σε συνδυασμό με την ανάπτυξη του Διαδικτ... more Η ανάπτυξη των Τεχνολογιών Πληροφορικής και Επικοινωνιών σε συνδυασμό με την ανάπτυξη του Διαδικτύου έχει αυξήσει σημαντικά την ποσότητα δεδομένων σε πληθώρα πεδίων, συμπεριλαμβανομένης της Εκπαίδευσης. Η υγειονομική κρίση και η εκτεταμένη χρήση διαδικασιών Εκπαίδευσης από Απόσταση οδήγησε σε περαιτέρω αύξηση εκπαιδευτικών δεδομένων. Ο τομέας της Εκπαίδευσης μπορεί να ωφεληθεί, με την αξιοποίηση των δεδομένων που παράγονται από τη διαδικασία της τηλεκπαίδευσης, μετά από επεξεργασία τους με χρήση τεχνικών και εργαλείων εξόρυξης δεδομένων. Έτσι, δίνεται η δυνατότητα για εξαγωγή πολύτιμης γνώσης που μπορεί να συμβάλει στην λήψη αποφάσεων σχετικά με την εκπαιδευτική διαδικασία. Στην παρούσα εργασία, χρησιμοποιήθηκε ανωνυμοποιημένα δεδομένα από το προπτυχιακό μάθημα «Βάσεις Δεδομένων» του τμήματος Μηχανικών Πληροφορικής και Υπολογιστών του Πανεπιστημίου Δυτικής Αττικής. Η επεξεργασία των δεδομένων έγινε με το εργαλείο RapidMiner Studio και χρησιμοποιήθηκαν τεχνικές clustering και δέντρων...
dissemination of electronic healthcare records in distributed wireless environments
ToA estimation system for efficient cycling in Smart Cities
Proceedings of the 21st Pan-Hellenic Conference on Informatics, 2017
Smartphone sensing enables efficient Time-of-Arrival (ToA) estimation of moving cyclists towards ... more Smartphone sensing enables efficient Time-of-Arrival (ToA) estimation of moving cyclists towards traffic lights in a Smart City. GPS sensors locate the actual position of cyclists on their way to traffic lights. Since the constant use of GPS sensors drain the battery of the smartphones there is a need of efficient energy consumption techniques. In this paper we propose a velocity based ToA estimation system to face the GPS energy consumption inefficiency. In addition we aim to enable efficient cycling in Smart Cities by turning traffic lights to green proactively thus achieving overall citizen's wellbeing.
Intelligent Tutoring Systems, 2020
In this paper, we propose and discuss a conceptual framework, based on knowledge management and s... more In this paper, we propose and discuss a conceptual framework, based on knowledge management and selective literature review, for enhancing the contribution of students in the Institutional Quality Assurance processes (IQA). The framework is related to the design of a mechanism for IQA, which ensures the improvement of the assessment and uses various methods, indicators, and criteria of auditing. The mechanism is supported by an integrated system for the evaluation of the learning practice, and it also includes a subsystem for evaluation based on Social Networking, a subsystem of learning analytics, and a subsystem for increasing the commitment to the IQA.
Journal Scientific and Technical Of Information Technologies, Mechanics and Optics, 2014
Humans are considered to reason and act rationally and that is believed to be their fundamental d... more Humans are considered to reason and act rationally and that is believed to be their fundamental difference from the rest of the living entities. Furthermore, modern approaches in the science of psychology underline that humans as a thinking creatures are also sentimental and emotional organisms. There are fifteen universal extended emotions plus neutral emotion: hot anger, cold anger, panic, fear, anxiety, despair, sadness, elation, happiness, interest, boredom, shame, pride, disgust, contempt and neutral position. The scope of the current research is to understand the emotional state of a human being by capturing the speech utterances that one uses during a common conversation. It is proved that having enough acoustic evidence available the emotional state of a person can be classified by a set of majority voting classifiers. The proposed set of classifiers is based on three main classifiers: kNN, C4.5 and SVM RBF Kernel. This set achieves better performance than each basic classifier taken separately. It is compared with two other sets of classifiers: one-against-all (OAA) multiclass SVM with Hybrid kernels and the set of classifiers which consists of the following two basic classifiers: C5.0 and Neural Network. The proposed variant achieves better performance than the other two sets of classifiers. The paper deals with emotion classification by a set of majority voting classifiers that combines three certain types of basic classifiers with low computational complexity. The basic classifiers stem from different theoretical background in order to avoid bias and redundancy which gives the proposed set of classifiers the ability to generalize in the emotion domain space.
Opinion mining using an LVQ neural network
Proceedings of the 21st Pan-Hellenic Conference on Informatics, 2017
Due to the increased use of social media in the past few years, a large volume of data has been a... more Due to the increased use of social media in the past few years, a large volume of data has been accumulated which contains human sentiments and opinions. The field that deals with the automated extraction of opinions is named opinion mining. In this paper, we evaluate the performance of an LVQ neural network on document level analysis using a benchmark movie review dataset. Document-level opinion mining aims at classifying a text, usually as positive or negative based on its overall sentiment. In order to reduce the dimensions of the reviews' vector representations, we use the feature selection method Information Gain. We use an exhaustive grid search for hyperparameter tuning and two methods for performance evaluation: a nested cross validation and a non-nested 10-fold cross validation. We study the performance of our model for different numbers of selected features by Information-Gain.
Uploads
Papers by CHRISTOS SKOURLAS