Skip to main content

Über dieses Buch

This book constitutes the refereed proceedings of two International Workshops held as parallel events of the 15th IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations, AIAI 2019, in Hersonissos, Crete, Greece, in May 2019: the 8th Mining Humanistic Data Workshop, MHDW 2019, and the 4th Workshop on 5G-Putting Intelligence to the Network Edge, 5G-PINE 2019.

The 6 full papers and 4 short papers presented at MHDW 2019 were carefully reviewed and selected from 13 submissions; out of the 14 papers submitted to 5G-PINE 2019, 6 were accepted as full papers and 1 as short paper. The MHDW papers focus on the application of innovative as well as existing data matching, fusion and mining and knowledge discovery and management techniques (such as decision rules, decision trees, association rules, ontologies and alignments, clustering, filtering, learning, classifier systems, neural networks, support vector machines, preprocessing, post processing, feature selection, visualization techniques) to data derived from all areas of humanistic sciences, e.g., linguistic, historical, behavioral, psychological, artistic, musical, educational, social, and ubiquitous computing and bioinformatics. The papers presented at 5G-PINE focus on several innovative findings coming directly from modern European research in the area of modern 5G telecommunications infrastructures and related innovative services and cover a wide variety of technical and business aspects promoting options for growth and development.



4th Workshop on “5G-Putting Intelligence to the Network Edge” (5G-PINE 2019)


A Cloud-Based Architecture for Video Services in Crowd Events

In this paper, we discuss a use case focused on crowd events in the context of the original 5G ESSENCE project [1], emphasizing on the provision of video services to the involved end-users. In particular, we investigate the core architectural components, as well as the cloud testbed for the deployment of our use case.
Alexandros Kostopoulos, Ioannis Chochliouros, Evangelos Sfakianakis, Daniele Munaretto, Claus Keuker

A New Self-planning Methodology Based on Signal Quality and User Traffic in Wi-Fi Networks

Wi-Fi networks have become one of the most popular technologies for the provisioning of multimedia services. Due to the exponential increase in the number of Access Points (AP) in these networks, the automation of the planning, configuration, optimization and management tasks has become of prime importance. The efficiency of these automated processes can be improved with the inclusion of data analytics mechanisms able to process the large amount of data that can be collected from Wi-Fi networks by powerful monitoring systems. This paper presents a new self-planning methodology that collects historical network measurements and extracts knowledge about user signal quality and traffic demands to determine adequate AP relocations. The performance of the proposed AP relocation methodology based on a genetic algorithm is validated in a real Wi-Fi network. The proposed approach can be easily adapted to other contexts such as small cell networks.
Juan Sánchez-González, Jordi Pérez-Romero, Oriol Sallent

Enhanced Mobile Broadband as Enabler for 5G: Actions from the Framework of the 5G-DRIVE Project

In the new fascinating era of 5G, new communication requirements set diverse challenges upon existing networks, both in terms of technologies and business models. One among the essential categories of the innovative 5G mobile network services is the enhanced Mobile Broadband (eMBB), mainly aiming to fulfill users’ demand for an increasingly digital lifestyle and focusing upon facilities that implicate high requirements for bandwidth. In this paper we have discussed eMBB as the first commercial use of the 5G technology. Then, we have focused upon the original context of the 5G-DRIVE research project between the EU and China, and we have identified essential features of the respective eMBB trials, constituting one of the corresponding core activities. In addition, we have discussed proposed scenarios and KPIs for assessing the scheduled experimental work, based on similar findings from other research and/or standardization activities.
Ioannis P. Chochliouros, Na Yi, Anastasia S. Spiliopoulou, Alexandros Kostopoulos, Nathan Gomes, Uwe Herzog, Tao Chen, Athanassios Dardamanis, Olga Segou, Philippos Assimakopoulos, Klaus Moessner, Juha Jidbeck, Rahim Tafazolli, Latif Ladid

Inclusion of Telemetry and Data Analytics in the Context of the 5G ESSENCE Architectural Approach

The 5G will not only be a kind of progress of mobile broadband networks but will also create a set of novel and unique network and service capabilities, structuring a form of a sustainable and scalable technology. Based on the context of the on-going progress of the actual “5G-ESSENCE” EU-funded project and, in particular, upon its innovative architecture that combines a variety of features from network functions virtualisation (NFV), mobile-edge computing (MEC) and cognitive network management resulting in a pure software-driven environment in nature, we identify the importance of telemetry and analytics. These latter features are expected to play an important role in the 5G ecosystem, especially for the realisation and support of dynamic cognitive management of the 5G ESSENCE network architecture. The Cloud Enabled Small Cell Manager (CESCM) which is a “core” element of the corresponding 5G ESSENCE architectural framework encompasses telemetry and analytics as essential tools for automated and fine grained management of the network infrastructures. To this aim, we have proposed the inclusion of three distinct modules (telemetry, analytics and orchestration) to enhance the original 5G ESSENCE architecture.
Ioannis P. Chochliouros, Anastasia S. Spiliopoulou, Alexandros Kostopoulos, George Agapiou, Maria Belesioti, Evangelos Sfakianakis, Michail-Alexandros Kourtis, Mike Iosifidis, Marinos Agapiou, Pavlos Lazaridis

A Framework to Support the Role of Telecommunication Service Providers in Evolving 5G Business Models

5G networks will constitute a complete transformation in the ICT domain by enabling the deployment of vertical services within the network infrastructures, based on extensive use of network softwarization and programmability. This shift will trigger and facilitate the transformation of existing stakeholders’ roles, as well as the interactions between multiple stakeholders from the traditionally separated markets. The 5G-PPP project MATILDA aims at delivering a holistic 5G end-to-end services operational framework, including 5G-ready applications lifecycle management from development to deployment over 5G network infrastructures. This paper aims at providing a refined and extended vision of the 5G business roles and their interactions and based on these at defining business applicability of the MATILDA project, with special focus on the project’s value proposition addressing the Telecommunication Service Providers.
Ioanna Mesogiti, Eleni Theodoropoulou, George Lyberopoulos, Fotini Setaki, Aurora Ramos, Panagiotis Gouvas, Anastasios Zafeiropoulos, Roberto Bruschi

Orchestration of Mission-Critical Services over an NFV Architecture

In the race towards 5G, NFV (Network Functions Virtualization) arises as one of the enabler technologies. The intelligent orchestration of the network becomes a key element to achieve the demanded network slicing for an efficient allocation of the available shared virtualised resources.
In this paper we propose an intelligent orchestration process of mission critical services over an NFV architecture. Mission critical services have tight requirements in terms of latency and high-availability that must be met in an end-to-end basis. Our proposal includes a monitoring system that collects performance data from the VNF (Virtual Network Function) instances in order to feed the decision-making process of the orchestrator and then elastically assign resources to the network service.
The software components that compose our deployment are presented as well as the validation scenario in which the features of the test-bed are exposed.
Aitor Sanchoyerto, Ruben Solozabal, Bego Blanco, Elisa Jimeno, Endika Aldecoa, Estrella Basurto, Fidel Liberal

Testbeds for the Implementation of 5G in the European Union: The Innovative Case of the 5G-DRIVE Project

An essential part of the actual EU policy towards promoting and validating 5G applications and of related solutions is via the establishment of an explicit plan and of a detailed roadmap for trials, tests and experimental activities though dedicated testbeds, in parallel with the current research and development activities coming from the 5G-PPP framework. The present paper discusses the fundamental role of the proposed trials’ initiatives within the broader European framework for the establishment and the promotion of 5G and also analyses the corresponding streams as indispensable parts of the 5G-PPP context, aiming to support innovation and growth. In addition, as part of the broader initiative for trial actions we identify the case of the 5G-DRIVE project that aims to realise 5G deployment scenarios (i.e., enhanced Mobile Broadband and Vehicle-to-Everything communications), between the EU and China, by discussing the fundamental features of the respective trials sites.
Ioannis P. Chochliouros, Anastasia S. Spiliopoulou, Alexandros Kostopoulos, Eirini Vasilaki, Uwe Herzog, Athanassios Dardamanis, Tao Chen, Latif Ladid, Marinos Agapiou

A Techno-Economic Analysis of Employing a Central Coordinator Entity in 5G Networks

This research work describes the role of the Central Controller and Coordinator (C3) entity and its potential techno-economic gain when implemented in the upcoming 5G networks. We investigate how viable could be for a C3 Producer and for a cellular network Operator to produce and implement respectively the C3 entity in its network. The performance of techno-economic analysis is estimated by considering various key parameters and some useful conclusions are drawn.
Christos Tsirakis, Mariana Goldhamer, Panagiotis Matzoros, George Agapiou, Dimitris Varoutas, Marinos Agapiou

Trustworthy AI for 5G: Telco Experience and Impact in the 5G ESSENCE

This paper discusses how it is possible to implement a general model to cope with the network and service management when the AI (Artificial Intelligence) capabilities are integrated in a trustworthy AI for 5G solutions framework. We discuss the general regulatory considerations related to the AI introduction in the 5G network management as this has been examined as part of the H2020 Project 5GESSENCE.
Maria Rita Spada, Alessandro Vincentini

Regulatory Considerations in the 5G Era: The 5GCity Neutral Host Case

The need for densification in 5G networks will necessitate the installation of large amounts of small cells in dense urban environments in order to meet future demands for capacity. The Neutral Host model has been proposed as a potential solution to this problem. This paper discusses the general regulatory considerations of 5G networks and specific considerations that relate to the Neutral Host model as this has been examined as part of the H2020 Project 5GCity.
Ioannis Neokosmidis, Theodoros Rokkas, Dimitrios Xydias, Maria Rita Spada

8th Mining Humanistic Data Workshop (MHDW 2019)


An Approach for Domain-Specific Design Pattern Identification Based on Domain Ontology

In this work, we present an approach for supporting the identification of domain-specific design patterns based on domain’s ontology, since the latter encapsulates the knowledge about the problem domain. More specifically, the proposed approach automatically analyzes the designs of a collection of domain-specific websites in terms of all the recurrent patterns occurring among them, both in the organization of their content and the front-end interface of their pages, resulting in a set of reusable design solutions which are commonly used in them by designers as building blocks for addressing typical domain problems. Then, evaluation is performed according to a number of inspection steps. At a first level, the recurrent patterns occurring at content organization, i.e., the common configurations of domain concepts occurring among website pages are evaluated by matching them against the domain’s ontology and selecting the ones which are in alignment with the domain’s context. At a second level, the recurrent patterns occurring at front-end organization (i.e., the common configurations of front-end design elements) are evaluated towards their consistent and effective use in designs of the collected websites. Finally, the approach categorizes the various reusable design solutions and recommends the ones with the best evaluation results as candidate domain-specific design patterns.
Vassiliki Gkantouna, Vaios Papaioannou, Giannis Tzimas, Zlatan Sabic

Pre-processing Framework for Twitter Sentiment Classification

Twitter Sentiment Classification is undergoing great appeal from the research community; also, user posts and opinions are producing very interesting conclusions and information. In the context of this paper, a pre-processing tool was developed in Python language. This tool processes text and natural language data intending to remove wrong values and noise. The main reason for developing such a tool is to achieve sentiment analysis in an optimum and efficient way. The most remarkable characteristic is considered the use of emojis and emoticons in the sentiment analysis field. Moreover, supervised machine learning techniques were utilized for the analysis of users’ posts. Through our experiments, the performance of the involved classifiers, namely Naive Bayes and SVM, under specific parameters such as the size of the training data, the employed methods for feature selection (unigrams, bigrams and trigrams) are evaluated. Finally, the performance was assessed based on independent datasets through the application of k-fold cross validation.
Elias Dritsas, Gerasimos Vonitsanos, Ioannis E. Livieris, Andreas Kanavos, Aristidis Ilias, Christos Makris, Athanasios Tsakalidis

Computing Long Sequences of Consecutive Fibonacci Integers with TensorFlow

Fibonacci numbers appear in numerous engineering and computing applications including population growth models, software engineering, task management, and data structure analysis. This mandates a computationally efficient way for generating a long sequence of successive Fibonacci integers. With the advent of GPU computing and the associated specialized tools, this task is greatly facilitated by harnessing the potential of parallel computing. This work presents two alternative parallel Fibonacci generators implemented in TensorFlow, one based on the well-known recurrence equation generating the Fibonacci sequence and one expressed on inherent linear algebraic properties of Fibonacci numbers. Additionally, the question of using lookup tables in conjunction with spline interpolation or direct computation within a parallel context for the computation of the powers of known quantities is explored. Although both parallel generators outperform the baseline serial implementation in terms of wallclock time and FLOPS, there is no clear winner between them as the results rely on the number of integers generated. Additionally, replacing computations with a lookup table degrades performance, which can be attributed to the frequent access to the shared memory.
Georgios Drakopoulos, Xenophon Liapakis, Evaggelos Spyrou, Giannis Tzimas, Phivos Mylonas, Spyros Sioutas

Employing Constrained Neural Networks for Forecasting New Product’s Sales Increase

An intelligent sales forecasting system is considered a rather significant objective in the food industry, since a reasonably accurate prediction has the possibility of gaining significant profits and better stock management. Many food companies and restaurants strongly rely on their previous data history for predicting future trends in their business operations and strategies. Undoubtedly, the area of retail food analysis has been dramatically changed from a rather qualitative science based on subjective or judgemental assessments to a more quantitative science which is also based on knowledge extraction from databases. In this work, we evaluate the performance of weight-constrained neural networks for forecasting new product’s sales increase. These new prediction models are characterized by the application of conditions on the weights of the network in the form of box-constraints, during the training process. The preliminary numerical experiments demonstrate the classification efficiency of weight-constrained neural networks in terms of accuracy, compared to state-of-the-art machine learning prediction models.
Ioannis E. Livieris, Niki Kiriakidou, Andreas Kanavos, Gerasimos Vonitsanos, Vassilis Tampakas

Language Processing for Predicting Suicidal Tendencies: A Case Study in Greek Poetry

Natural language processing has previously been used with fairly high success to predict a writer’s likelihood of committing suicide, using a wide variety of text types, including suicide notes, micro-blog posts, lyrics and even poems. In this study, we extend work done in previous research to a language that has not been tackled before in this setting, namely Greek. A set of language-dependent (but easily portable across languages) and language-independent linguistic features is proposed to represent the poems of 13 Greek poets of the 20th century. Prediction experiments resulted in an overall classification rate of 84.5% with the C4.5 algorithm, after having tested multiple machine-learning algorithms. These results differ significantly from previous research, as some features investigated did not play as significant a role as was expected. This kind of task presents multiple difficulties, especially for a language where no previous research has been conducted. Therefore, a significant part of the annotation process was performed manually, which likely explains the somewhat higher classification rates compared to previous efforts.
Alexandros Dimitrios Zervopoulos, Evangelos Geramanis, Alexandros Toulakis, Asterios Papamichail, Dimitrios Triantafylloy, Theofanis Tasoulas, Katia Kermanidis

Recognition of Urban Sound Events Using Deep Context-Aware Feature Extractors and Handcrafted Features

This paper proposes a method for recognizing audio events in urban environments that combines handcrafted audio features with a deep learning architectural scheme (Convolutional Neural Networks, CNNs), which has been trained to distinguish between different audio context classes. The core idea is to use the CNNs as a method to extract context-aware deep audio features that can offer supplementary feature representations to any soundscape analysis classification task. Towards this end, the CNN is trained on a database of audio samples which are annotated in terms of their respective “scene” (e.g. train, street, park), and then it is combined with handcrafted audio features in an early fusion approach, in order to recognize the audio event of an unknown audio recording. Detailed experimentation proves that the proposed context-aware deep learning scheme, when combined with the typical handcrafted features, leads to a significant performance boosting in terms of classification accuracy. The main contribution of this work is the demonstration that transferring audio contextual knowledge using CNNs as feature extractors can significantly improve the performance of the audio classifier, without need for CNN training (a rather demanding process that requires huge datasets and complex data augmentation procedures).
Theodore Giannakopoulos, Evaggelos Spyrou, Stavros J. Perantonis

Studying the Spatialities of Short-Term Rentals’ Sprawl in the Urban Fabric: The Case of Airbnb in Athens, Greece

This work constitutes a theoretically-informed empirical analysis of the spatial characteristics of the short-term rentals’ market and explores their linkage with shifts in the wider housing market within the context of a south-eastern EU metropolis. The same research objective has been pursued for a variety of international paradigms; however, to the best of our knowledge, there has not been a thorough and systematic study for Athens and its neighborhoods. With a theoretical framework that draws insight from the political-economic views of Critical Geography, this work departs from an assessment of Airbnb listings, and proceeds inquiring the expansion of the phenomenon with respect to the rates of long-term rent levels in the neighborhoods of Central Athens, utilizing relevant data. The geographical framework covers the City of Athens as a whole, an area undergoing profound transformations in recent years, stemming from diverse factors that render the city one of the most dynamic destinations of urban tourism and speculative land investment. The analysis reveals a prominent expansion of the short-term rental phenomenon across the urban fabric, especially taking ground in hitherto underexploited areas. This expansion is multifactorial, asynchronous and exhibits signs of positive relation with the long-term rentals shifts; Airbnb not only affects already gentrifying neighborhoods, but contributes to a housing market disruption in non-dynamic residential areas.
Konstantinos Gourzis, Georgios Alexandridis, Stelios Gialis, George Caridakis

A Brief Overview of Dead-Zone Pattern Matching Algorithms

Within the last decades, the dead-zone algorithms have emerged as being highly performant on certain types of data. Such algorithms solve the keyword exact matching problem over strings, though extensions to trees and two-dimensional data have also been devised. In this short paper, we give an overview of such algorithms.
Miznah Alshammary, Mai Alzamel, Costas Iliopoulos, Richard E. Watson, Bruce W. Watson

On the Cyclic Regularities of Strings

Regularities in strings are often related to periods and covers, which have extensively been studied, and algorithms for their efficient computation have broad application. In this paper we concentrate on computing cyclic regularities of strings, in particular, we propose several efficient algorithms for computing: (i) cyclic periodicity; (ii) all cyclic periodicity; (iii) maximal local cyclic periodicity; (iv) cyclic covers.
Oluwole Ajala, Miznah Alshammary, Mai Alzamel, Jia Gao, Costas Iliopoulos, Jakub Radoszewski, Wojciech Rytter, Bruce Watson

Paid Crowdsourcing, Low Income Contributors, and Subjectivity

Scientific projects that require human computation often resort to crowdsourcing. Interested individuals can contribute to a crowdsourcing task, essentially contributing towards the project’s goals. To motivate participation and engagement, scientists use a variety of reward mechanisms. The most common motivation, and the one that yields the fastest results, is monetary rewards. By using monetary, scientists address a wider audience to participate in the task. As the payment is below minimum wage for developed economies, users from developing countries are more eager to participate. In subjective tasks, or tasks that cannot be validated through a right or wrong type of validation, monetary incentives could contrast with the much needed quality of submissions. We perform a subjective crowdsourcing task, emotion annotation, and compare the quality of the answers from contributors of varying income levels, based on the Gross Domestic Product. The results indicate a different contribution process between contributors from varying GDP regions. Low income contributors, possibly driven by the monetary incentive, submit low quality answers at a higher pace, while high income contributors provide diverse answers at a slower pace.
Giannis Haralabopoulos, Christian Wagner, Derek McAuley, Ioannis Anagnostopoulos

Predicting Secondary Structure for Human Proteins Based on Chou-Fasman Method

Proteins are constructed by the combination of a different number of amino acids and thus, have a different structure and folding depending on chemical reactions and other aspects. The protein folding prediction can help in many healthcare scenarios to foretell and prevent diseases. The different elements that form a protein give the secondary structure. One of the most common algorithms used for secondary structure prediction constitutes the Chou-Fasman method. This technique divides and in following analyses each amino acid in three different elements, which are -helices, -sheets and turns based on already known protein structures. Its aim is to predict the probability for which each of these elements will be formed. In this paper, we have used Chou-Fasman algorithm for extracting the probabilities of a series of amino acids in FASTA format. We make an analysis given all probabilities for any length of a human protein without any restriction as other existing tools.
Fotios Kounelis, Andreas Kanavos, Ioannis E. Livieris, Gerasimos Vonitsanos, Panagiotis Pintelas


Weitere Informationen

Premium Partner