Traži  English (United States) Hrvatski (Hrvatska)

inovativno promotivno partnerstvoUmjetnom inteligencijom prema EU višejezičnosti

Tehničko kosponzorstvo

 
MIPRO 2019 - 42. međunarodni skup

miproBIS - Poslovni inteligentni sustavi

srijeda, 22.5.2019 14:00 - 19:00, Nava 1, Hotel Admiral, Opatija




Prezentirani radovi na engleskom jeziku bit će poslani za uključenje u digitalnu bazu IEEE Xplore.

Program događaja
srijeda, 22.5.2019 14:00 - 19:00,
Nava 1, Hotel Admiral, Opatija
Radovi 
1.D. Orescanin (Poslovna inteligencija, Zagreb, Croatia), R. Tan, J. Ao (North Carolina State University, Raleigh, United States)
Conceptual Framework for Entity Integration from Multiple Data Sources 
Entity matching and entity resolution are becoming more important disciplines in data management over time, based on increasing number of data sources that should be addressed in economy that is undergoing digital transformation process, growing data volumes and increasing requirements related to data privacy. Entity resolution is element of larger entity integration process that include data acquisition, data profiling, data cleansing, schema alignment and mapping, data matching and data merge (fusion). This paper is a part of the work in progress aimed to define framework that will automate some steps entity integration process and try to reduce requirements for human influence in the process. In this paper focus will be on conceptual process definition and use of existing open source solutions for entity integration process automation
2.M. Francia, E. Gallinucci, F. Vitali (University of Bologna, Cesena, Italy)
Map-Matching on Big Data: A Distributed and Efficient Algorithm with a Hidden Markov Model 
In urban mobility, map-matching aims to project GPS points generated by moving objects onto the road segments representing the actual object positions. Up to now, map-matching has found interesting applications in traffic analysis, frequent path extraction, and location prediction. However, state-of-art implementations of map-matching algorithms are either private, sequential or inefficient. In this paper, we propose an extension of an existing serial algorithm of known efficiency by reformulating it in a parallel way, in order to achieve great scalability on real big data scenarios. Furthermore, we enhance the robustness of the algorithm by introducing a smart strategy based on a first order Hidden Markov Model to avoid gaps in the matched road segments, i.e., a problem that may occur under sparse GPS sampling or in urban areas, where road network is highly fragmented. Our implementation is based on Apache Spark and is publicly available on Github. The implementation is tested against a dataset with 300 million GPS points in Milan.
3.V. Nicheporchuk, T. Penkova, I. Gryazin (Institute of Computational Modeling of the Siberian Branch of the Russian Academy of Science Krasnoy, Krasnoyarsk, Russian Federation)
Structuring the Information Resources for Intelligent Ecosystem Monitoring System Based on Camera Traps 
Monitoring of animals in the wild using camera traps is one of the promising ways for monitoring animal behavior and estimation of dynamic of changes in natural systems over the large areas. Modern tools of video registration function 24/7/365, some of them produce a set of still images of moving object (animal, bird, or human) through 5-8 seconds, while other devices deliver a short movie. The installation of camera traps in a predefined place, such as animal shelters, trails, watering places, provides unique information that is not available by applying other methods of monitoring and control. Present a lot of methods and algorithms have been developed that allow us to recognize the natural images taking into account the variability of shooting conditions, illumination, location and posture of the animal, as well as a complex textural background, representing various states of the forest, steppe, mountain and tundra landscapes. This paper presents the results of data analysis and systematization from images obtained on the territory of wildlife sanctuaries in the Krasnoyarsk region. The authors proposed the content and structure of data warehouse for intellectual ecosystem monitoring system. The structure of specialized data warehouse combines historical, operational and reference data. To provide the application of the modern data processing techniques the data warehouse includes three principal layers: layer of stationary storage that contains the main part of monitoring data and consists of facts tables and dimensions tables; layer of pre-loading processing that is necessary for loading and preliminary processing of monitoring data and contains data sources, import and export procedures and temporary tables; layer of analytical objects that contains the analytical models which are used for on-line analytical processing and data mining, aggregates tables and analytical reports. Finally, the authors determine the requirements for representation of data analysis results to estimate the state of populations, ecosystems and possible threats in the natural conservation areas
4.M. Puh, L. Brkić (Fakultet elektrotehnike i računarstva, Zagreb, Croatia)
Detecting Credit Card Fraud Using Selected Machine Learning Algorithms 
Due to the immense growth of e-commerce and increased online based payment possibilities, credit card fraud has become deeply relevant global issue. Recently, there has been major interest for applying machine learning algorithms as data mining technique for credit card fraud detection. However, number of challenges appear, such as lack of publicly available data sets, highly imbalanced class sizes, variant fraudulent behaviour etc. In this paper we compare performance of three machine learning algorithms: Random forest, Support Vector Machine and Logistic regression in detecting fraud on real-life data containing credit card transactions. To mitigate imbalanced class sizes, we use SMOTE sampling method. The problem of ever-changing fraud patterns is considered with employing incremental learning of selected ML algorithms in experiments. The performance of the techniques is evaluated based on commonly accepted metric: precision and recall.
5.R. Delina (Technical University, Košice, Slovakia), L. Senderáková, R. Olejárová, M. Macík (Technical University , Košice, Slovakia)
An Empirical Study on the Impact of New Participants on Final Price Savings and Process of Electronic Reverse Auctions 
The electronic reverse auction as a popular procurement mechanism, whose use can secure higher objectiveness of prices or utility of products and services across different industries, is the subject of research already for several years, however, the range of empirical studies is still lagging behind. The current paper extends the empirical findings on the subject of the electronic reverse auction. The aim of our study is to examine the impact of including new participants in the auction process. The study is based on the database of 4,442 real auction data cases of various companies including both public and private entities and different types of auction collected by the most significant reverse auction solution provider in central Europe. The research results confirmed the impact of new participants in the electronic reverse auction. According to our analysis, we have identified that there is a positive effect of new participants on final price savings.
6.T. Bertoncel, M. Meško (University of Primorska, Faculty of management, Koper, Slovenia), M. Pejić Bach (Ekonomski fakultet Zagreb, ZAGREB, Croatia)
Big Data for Smart Factories: A Bibliometric Analysis 
The use of big data is rapidly growing in recent years, as a result of access to huge amounts of data, which can be obtained from various sources. Big data has played a significant role in all sectors, for example, the manufacturing sector, where it plays a very important role due to its ability to provide economic growth. Several studies have been done regarding the use of big data at smart factories, for this reason, the main aim of our study was to conduct a bibliometric and phrase frequency analysis that covered the name of most frequent authors, journals and organizations, year and country of publication and the frequency with which words and phrases appeared in the literature. Using the keywords “big data” and “smart factory,” we searched bibliometric units from the Scopus database. We found out that research on big data and smart factories is increasing rapidly and is mainly focused on machine learning and cloud computing.
7.T. Zovak, A. Šarčević, M. Vranić, D. Pintar (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
Game-to-Game Prediction of NBA Players’ Points in Relation to Their Season Average 
NBA attracts a great deal of attention among sports analysts and sports betting institutions regarding the prediction of various outcomes of each game, together with the parameters which affect them. Performance of NBA players is influenced by many unknown and random factors, such as players’ psychological condition, social life and injuries. The stated factors hinder game-to-game predictions of players’ performance in relation to the expectations set by their past performances. Using statistics available to the public, a dataset was created, in which one instance represents one game of one player in a season. A comparison between points that a player has scored, and his current season average was done in order to label the instance in one of the two categories: ‘over’ or ‘under’. Using various statistical data concerning previous games of the season, a binary classifier was trained in order to distinguish between those categories for future games. The classifier showed to perform with an F1 score of 56.7%. Since sports betting institutions tend to give 50/50 odds of a player going ‘over’ or ‘under’ in relation to their season points per game, these results represent an improvement of 6.7%. Although top features are predominated by offensive statistics (e.g. how many minutes the player plays, how many shots he takes and how strong the offense of his team is), a newly generated feature, which represents tiredness of a player, has shown to be among top 15 informative features.
8.R. Delina, R. Olejárová, M. Macik, L. Senderáková (TUKE, Košice, Slovakia)
Analysis of the Simulation-Based Procurer Behavior Determinants 
Digital economy brought higher transparency of the markets where more non-price factors are important in the procurement decision-making of global competition. Procurers state that numerous researchers in marketing have recognized the critical roles of corporate reputation in customer’s buying behavior, which means corporate reputation plays a fundamental role. The process of the procurement forms a large part of both corporate and public activities and absorbs many financial and non-financial costs. To decrease those costs, it is essential to understand the dynamics of the procurer - supplier relationship. The purpose of this article is to identify main characteristics that include a price and non-price oriented determinants of the suppliers that are crucial for the procurers. The impact of these characteristics was determined by descriptive statistics, decision trees and correlation analysis over a database of simulations completed by professional procurers of the Slovak companies. The results of the analysis suggest the most influential analyzed parameter is the supplier’s price level, but procurers are also influenced by non-price factors as internal and external loyalty ratio to set the preferences about suppliers.
9.A. Tunjić (Multicom d.o.o., Zagreb, Croatia)
Automation of Data Lake Ingestion Process from Various Sources 
In a big data environment, it is often necessary to ingest data from different sources into a unique storage. Because of low memory price, system distribution and failure tolerance, that storage is typically HDFS. It enables users to manipulate data with different tools from the Hadoop ecosystem. The process of data ingestion seems simple, however, because sources can be different database systems, structured, semi-structured and unstructured data complicate the ingestion procedure. It is usually not enough to just store everything, data needs to be stored in such a way that enables users to quickly access and manipulate it. There are many ingestion-specific solutions in the big data ecosystem. This paper will describe an implemented system for data ingestion from MSSQL, MySQL and Postgres into a Hive database. It starts with creating tables with corresponding metadata, continues with the ingestion process and finally, how the process is automated. Implementation of Sqoop as an open-source tool and Hue, a web user interface from Cloudera will be described.
10.M. Matetić (Odjel za Informatiku, Sveučilište u Rijeci, Rijeka, Croatia)
Mining Learning Management System Data Using Interpretable Neural Networks 
The paper presents the work on data analysis of LMS data related to the course Programming 2, one of the introductory courses at the first year of the study of informatics at the Department of Informatics at the University of Rijeka. In order to improve the course, the analysis of data from the LMS system is carried out with an emphasis on some additional activities that are not scored, such as watching video lectures and self assessments. We are interested that these activities have positive impact on the student success and the data analysis can very objectively evaluate their meaning, the effect of improvement and their impact on the learning process. The paper presents experiences in the design of such additional activities and the discovery of knowledge about the process of learning using batch data analysis performed by Artificial Neural Networks. We also discuss the use of Recurrent Neural Network in order to analyze LMS time-series data and largely depends on how the course activities are planned and shaped.
11.V. Banović, I. Sorić (Poslovna inteligencija d.o.o., Zagreb, Croatia)
General Solution Framework for Management and Monitoring of Cross-Platform Data Processes 
The purpose of this paper is a description of a proof-of-concept implementation for running and managing cross-platform data processes. The term “data process” in this context represents a variety of data management processes being executed by any of the common data integration tools and business intelligence processes ran through BI tools. Besides the common data management process workflow, custom data processing can be achieved by executing scenario-specific scripts, such as Big Data processing being achieved by running Spark or Hadoop jobs from inside the platform. The framework is implemented as a web application with a separate service which handles job execution and scheduling. Besides that, the service addresses the question of achieving parallelism, performance optimization and dependency management including the automatic discovery of different dependencies. The user interface offers monitoring of process execution through a graph representation of the execution plan or a regular table visualization. This paper will explain the details of the developed concept and the architecture of our technology agnostic implementation.
12.A. Kutnjak, I. Pihir, M. Tomičić Furjan (Faculty of Organization and Informatics, University of Zagreb, Varaždin, Croatia)
Digital Transformation Case Studies Across Industries – Literature Review  
Digital transformation (DT) has been introduced in all industries for improving the way of running business. Industries have embraced technology in different extents, as not all technologies can be equally implemented even in semantically similar processes. This can be explained by the fact that the hype cycle of emerging technologies differently impact some industries in time between the concrete technology has been discovered, used for the first time and time when technology reach his full potential of use in real processes so called maturity. This paper use a literature review on Digital transformation papers that use case study method in their researches and compare digital transformation case studies across different industries. Paper analyse case studies to bring some light to real digital transformation processes and its success in practice. Analysed papers are selected from Web of Science database on content, industry and number of papers.
13.T. Mladenova (University of Ruse "Angel Kanchev", Ruse, Bulgaria)
A Project Managment System for Time Planning and Resources Allocation 
When comes to the efficient management and planning of a company’s resources and workload many modern businesses are facing a serious problem which can be the cause for their failure. Lack of planning is certain to result in shortages, reduced effectiveness and delays. This paper presents two types of analyses - an overview of existing solutions and an organization survey. Several Enterprise Resource Planning systems are reviewed and their positives and negatives are compared and rated by a set of criteria. The survey of the work processes of a small organization among with interviews with employees gives an insight of the problems and needs of a small-to-medium size businesses. The two sets of research complete the functional requirements for a project management system that can allocate resources more effectively. A web-based solution is proposed and is viewed through the prism of the whole business process and not just as an independent unit, although the presented application can be either a stand-alone system or a module of an ERP system with the project management being an operative module. Experimentally the project management system is implemented in the surveyed company replacing an existing PMS systems, the results of an additional survey are presented and alyzed as well.
14.L. Hrustek, M. Tomičić Furjan, I. Pihir (Fakultet organizacije i informatike, Sveučilište u Zagrebu, Varaždin, Croatia)
Influence of Digital Transformation Drivers on Business Model Creation 
Digital transformation (DTX) can be seen as an approach which assumes a radical change of doing business, driven highly by todays necessarity to change and adapt to the digital age. This change is mostly made by creating new business models, which define the way how an organization provides value to their customers. The drivers of DTX define ideas for organizational innovations, and the can arise from trends in the organizational environment or from innovation ideas within the organization itself. Either way, the desire to change can be classified as customer driven - when the change is a result of an adoption to new customer needs, technology driven - when a new technology has become a standard within the industry the business acts in or organizational development driven, when the need for digital improvement of work is defined as a strategic goal that has to be achieved. In this paper we define and analyse the drivers of DTX and explore their influence on the creation of new business models.
15.D. Etinger, S. Simić, L. Buljubašić (Juraj Dobrila University of Pula, Pula, Croatia)
Automated Decision-Making with DMN: from Decision Trees to Decision Tables  
Recent advances in artificial intelligence, especially the subfield of machine learning, is commonly cited as one of the driving forces for digital transformation and innovative business models. Ongoing research is focusing on embedding solutions based on machine learning into business processes which are commonly modelled using the BPMN standard. The Object Management Group has recently adopted the Decision Model and Notation standard. By using the Decision Model and Notation (DMN) it is possible to replace multiple decision points embedded in business processes. The purpose of this research is to provide a method to derive DMN decision tables from the corresponding model generated by the machine learning algorithms. The concept is being developed using the Python machine learning library scikit-learn and Camunda BPM. We propose a tool that integrates decision tree models with DMN tables. This approach automates the process of converting machine learning models into DMN tables which can be used alongside BPMN process models.
16.M. Nikitović (Visoka škola za informacijske tehnologije, Zagreb, Croatia), A. Mahmutović (Elektroprojekt, Zagreb, Croatia)
Hidden Costs of ERP Implementation 
It’s well known that ERP Implementation requires, above all, time and money. It can be said globally that these two elements represent a summary activities by many critical success factors (CSFs) of ERP implementation. Without diminishing the importance of time, cost component of ERP Implementation may be significantly increased by the hidden Costs of Implementation, which is in the phase of supply and contracting do not see as separate items. The paper presents research on hidden costs of ERP Implementation in small and medium-sized (SME) manufacturing companies in Croatia.

Osnovni podaci:
Voditelji:

Boris Vrdoljak (Croatia), Mirta Baranović (Croatia), Matteo Golfarelli (Italy), Mihaela Vranić (Croatia)

Programski odbor:

Marko Banek (Croatia), Mirta Baranović (Croatia), Bartosz Bebel (Poland), Ladjel Bellatreche (France), Ljiljana Brkić (Croatia), Alfredo Cuzzocrea (Italy), Todd Eavis (Canada), Enrico Gallinucci (Italy), Dragan Gamberger (Croatia), Paolo Garza (Italy), Matteo Golfarelli (Italy), Igor Mekterović (Croatia), Damir Pintar (Croatia), Rok Rupnik (Slovenia), Zoran Skočir (Croatia), A Min Tjoa (Austria), Mladen Varga (Croatia), Mihaela Vranić (Croatia), Boris Vrdoljak (Croatia) 

Prijava/Kotizacija:
PRIJAVA / KOTIZACIJE
CIJENA U EUR-ima
Do 6.5.2019.
Od 7.5.2019.
Članovi MIPRO i IEEE
200
230
Studenti (preddiplomski i diplomski studij) te nastavnici osnovnih i srednjih škola
120
140
Ostali
220
250

Popust se ne odnosi na studente doktorskog studija.

Kontakt:

Boris Vrdoljak
Fakultet elektrotehnike i računarstva
Unska 3
10000 Zagreb, Hrvatska

Tel.: +385 1 6129 532
Fax: +385 1 6129 915
E-mail: boris.vrdoljak@fer.hr 

Najbolji radovi bit će nagrađeni.
Prihvaćeni radovi bit će objavljeni u zborniku radova s ISSN brojem. Radovi napisani na engleskom jeziku i prezentirani na skupu bit će poslani za objavljivanje u bazi IEEE Xplore.



Mjesto održavanja:

Opatija, sa 170 godina dugom turističkom tradicijom, vodeće je ljetovalište na istočnoj strani Jadrana i jedno od najpoznatijih na Mediteranu. Ovaj grad aristokratske arhitekture i stila već 170 godina privlači svjetski poznate umjetnike, političare, kraljeve, znanstvenike, sportaše, ali i poslovne ljude, bankare, menadžere i sve kojima Opatija nudi svoje brojne sadržaje. 

Opatija svojim gostima nudi brojne komforne hotele, odlične restorane, zabavne sadržaje, umjetničke festivale, vrhunske koncerte ozbiljne i zabavne glazbe, uređene plaže i brojne bazene i sve što je potrebno za ugodan boravak gostiju različitih afiniteta. 

U novije doba Opatija je jedan od najpoznatijih kongresnih gradova na Mediteranu, posebno prepoznatljiva po međunarodnim ICT skupovima MIPRO koji se u njoj održavaju od 1979. godine i koji redovito okupljaju preko tisuću sudionika iz četrdesetak zemalja. Ovi skupovi Opatiju promoviraju u nezaobilazan tehnološki, poslovni, obrazovni i znanstveni centar jugoistočne Europe i Europske unije općenito.


Detaljnije informacije se mogu potražiti na www.opatija.hr i www.visitopatija.com.

Download
 
Novosti o događaju
Trenutno nema novosti
 
Suorganizatori - nasumično
FOI VaraždinIRB ZagrebHAKOMKončar Elektroindustrija ZagrebENT Zagreb