|
|
| Business Intelligence Systems - miproBIS 2010
May 27, 2010
9:00 - 13:00 h
Opatija, Hotel Admiral (Hall 3)
|
S. Pinjušić (Fakultet elektrotehnike i računarstva, Zagreb, Croatia), M. Vranić (Fakultet elektrotehnike i računarstva, Zagreb, Zagreb, Croatia)
Influence of distance measure choice on the results of hierarchical clustering 
Hierarchical clustering is used to assign observations into clusters which are further connected to form a hierarchical structure. Observations in the same cluster are close together according to the predetermined distance measure. In this paper different distance measures are explored and experimented with. Influence of different distance measures on results of hierarchical clustering is analyzed and elaborated. Data mining tool ‘Orange’ offers flexibility to reprogram some modules so it was used to process data and visualize results.
|
J. Vasilev, M. Kasheva, T. Atanasova, V. Kisimov, S. Sulova (Varna University of Economics, Varna, Bulgaria)
Bulgarian project “A methodology for design and implementation of a distributed business intelligent system for SMEs”
European and Bulgarian business climate requires the use of business intelligent systems (BIS) to support quick and adequate decisions in a competitive environment of the great dynamic European market. BIS are mainly used in huge corporations. SMEs also need BIS, but few SMEs use BIS. The paper presents a research supported by The National Science Fund of Bulgaria focused on the project “A methodology for design and implementation of a distributed business intelligent system for SMEs”. Results of a questionnaire about the need and usage of a distributed BIS among Bulgarian SMEs are presented. The offered distributed BIS meets completely the specification of SMEs. It will give enterprises a strong software solution for access to knowledge databases. It will extract expert opinion to help managerial decisions.
|
A. Karabegovic (Faculty of Electrical Engineering, Sarajevo, Bosnia and Herzegovina), M. Ponjavic (Gauss, Tuzla, Bosnia and Herzegovina)
Integration and Interoperability of Spatial Data in Spatial Decision Support System Environment
Nowadays, often picture of sharing data is: different organization on different levels hosts its own GIS. Each uses software and a data model which best meet their needs with centralized storage and metadata interoperability, using right tool for the job while eliminating complicated data transfers and multiple copies of the same data throughout the enterprise or department. Spatial Object Model enable using any combination of commercial or open-source GIS tools to work together with same database using triggers and storage processes. This paper demonstrates usage of IT infrastructure which is wide recognized by local governments or enterprises as platform for three-tier structure solution: spatial data server, application server and application client. Based on application server it is possible to produce different OGC Web Services for use spatial data in background application or expose via GIS portal. Any client can request the server if it accords with OGC specification. Objective of this paper is studies of very demanding integration and interoperability task from real-world: data server for storing terabytes of data, application server for creating web services which enable distributing this amount of data and simultaneously different GIS clients and WebGIS portal.
|
T. Piasevoli (SOFTPRO TETRAL d.o.o., Zagreb, Croatia)
Universal Calculations in MDX
Among many of its functions, MDX language has one special, but rarely used function - Axis() function. Through the use of that function it is possible to create calculations that do not reference any particular dimension/hierarchy in a cube but work correctly for any of them. These calculations can be used in any MDX query or report and are therefore named „universal“. The paper explains the principles behind these calculations and the techniques how to achieve them. It also explains how they work starting from a simple example and continuing with more advanced ones. Before the final conclusion is made, various approaches and techniques are being analyzed and compared.
|
A. Cvitaš (Oracle Hrvatska d.o.o., Zagreb, Croatia)
Information Extraction in Business Intelligence Systems
Business intelligence plays a key role in many business systems by enabling easier and quicker decision-making. The strength of business intelligence system is directly dependent of source data and it’s availability. This paper analyzes alternatives to usage of strictly structured data.
Text mining systems can function independently but can also be considered as just one step in larger information systems. Combination of business intelligence with information extraction methods can significantly improve data set used for analysis with the result of enlarged knowledge base for decision-making.
This paper gives an overview of many proposed sets of techniques for information extraction but focuses on machine learning as the most promising one. In order to minimize human effort necessary for machine learning but also to gain satisfying results the optimal choice would be weakly supervised machine learning techniques.
|
Z. Kazi (Technical faculty "Mihajlo Pupin", Zrenjanin, Serbia), B. Radulović, D. Radovanović, L. Kazi (Tehnički fakultet "Mihajlo Pupin", Zrenjanin, Serbia)
MOLAP Data Warehouse of a Software Products Servicing Call Center
This paper presents data warehouse system design on example of a software importer and distributor Call center. Data analyses of telephone requests for servicing software products at Call center has been implemented. Call center bussiness process has been described. Conceptual database model is presented, as well as aggregated data created with data warehouse. Multi-Dimensional On-Line Analytical Processing (MOLAP) cubes based on star and snowflake schema are implemented. Additional features of executing MDX queries and export of analytical data to spreadsheet application, for the purposes of reports in the form of pivot tables and charts, are developed.
|
I. Ninčević, M. Ćukušić, Ž. Garača (Faculty of Economics, Split, Croatia)
Mining Demographic Data with Decision Trees
In recent years, the industry has started to apply data mining methods to predict and prevent social problems. The purpose of this research is to discover and classify the impact of various factors on life expectancy at birth. To determine and present key indicators, demographic data mining was conducted based on available worldwide data from the period from 1990 to 2006. Extraction – transformation – loading (ETL) processes that preceded data mining are presented as well. Among other mining methods, decision tree method is selected as the optimal for the problem. The research was conducted using a freeware program for data-mining, Orange Canvas 2.0.
|
J. Vasilev, T. Atanasova, M. Kasheva, S. Sulova (Varna University of Economics, Varna, Bulgaria)
Analysis of the possible application of Data Mining, Text Mining, Web Mining in Business Intelligent Systems
The design and implementation of a contemporary Business Intelligent System (BIS) requires processing of all types of information – structured, semi-structured, unstructured and real time. The basic functionality of BIS is implemented in an enterprise local area network (LAN) or by web access to web sites. Business results are bettered by the application of the following technologies: Data Mining (DM), Text Mining (TM) and Web Mining (WM). As a result companies get competitive advantage. Applications of BIS in enterprises are structured in a functional matrix. Its dimensions are managerial duties and subsystems of a company. Several problems are analyzed. They may be solved by the application of technologies DM, TM and WM in BIS. This processes leads to innovative solutions.
|
M. Ljubić (Multicom d.o.o., Zagreb, Croatia)
Granularnost činjeničnih tablica u skladištu podataka. Gdje je granica denormalizacije?
Jedna od osnovnih zadaća u procesu modeliranja skladišta podataka jest određivanje granularnosti fakt tablica. Ona se u pravilu promatra u kontekstu dimenzijskih struktura, tj. željene razine analitike, a obično je vrlo blizu ili jednaka razini same transakcije koja predstavlja tzv. atomsku razinu podataka.
Teorija kaže da su fakt tablice obično visoko normalizirane (za razliku od dimenzijskih), a njihovo postojanje određeno je vrstom poslovnog događaja. No što kada jedan poslovni događaj uzrokuje više drugih, a svi su predmet iste poslovne analize i prikazuju se na istom izvješću?
Rad opisuje varijante modeliranja i performansne analize, te konačno razloge kreiranja denormaliziranih fakt tablica, uz prateću problematiku različitih granularnosti i dimenzionalnosti mjera, aditivnosti i semiaditivnosti, te primjenu kompleksne horizontalne sigurnosti na razini retka po nekonzistentnom kriteriju.
|
A. Fehratbegović (JP Elektroprivreda BiH, Sarajevo, Bosnia and Herzegovina)
Upravljanje poslovnim procesima uz pomoć poslovne inteligencije
Danas se poslovna inteligencija intenzivno koristi kao važan faktor za podršku poslovnom odlučivanju, a predviđanja analitičara su da će se trend korištenja poslovne inteligencije, kao i novčana ulaganja u istu u narednim godinama intenzivno povećavati.
U JP Elektroprivreda BiH su prepoznate prednosti korištenja izvještaja poslovne inteligencije, te je implementirano web bazirano izvještavanje koje omogućava menadžerima sa strateškog nivoa upravljanja kontrolu nad realizacijom operativnog plana ulaganja, kontrolu procesa planiranja, računovodstvenim knjiženjima, realizacijom projekata itd.
Cilj ovoga rada je ukazati na metodologiju koja se koristila za prikupljanje, analizu i izvještavanje, u cilju upravljanja poslovnim procesima u organizaciono i geografski distribuiranoj kompaniji kao što je to JP Elektroprivreda BiH. Također, cilj je da se predstave mogući pravci korištenja poslovne inteligencije u svrhu poboljšanja poslovanja kompanije u drugim segmentima. U tom smislu, potrebno je osvrnuti se i na činjenicu da JP Elektroprivreda prolazi kroz proces implementacije novog informacionog sistema (SAP), kao i proces pripajanja rudnika koji imaju vlastiti informacioni sistem.
|
C. Pansini (Fakultet elektrotehnike i računarstva, Zagreb, Croatia)
Korištenje alata otvorenog koda za poslovnu inteligenciju
Razvoj alata otvorenog koda sve je intenzivniji u raznim područjima informatike pa tako i u području poslovne inteligencije. Primjeni takvih alata još uvijek se pristupa s određenom nesigurnošću budući da se često nameće pitanje njihove kvalitete i upotrebljivosti u usporedbi s odgovarajućim komercijalnim rješenjima. Ovaj rad opisuje sustav skladištenja podataka čija je namjena analiza poslovanja videoteke. Skladište podataka izgrađeno je pomoću programskog alata Oracle Warehouse Builder, dok se za analizu podataka koriste odabrani alati otvorenog koda iz paketa JasperSoft BI Suite. Pomoću alata JasperAnalysis provedena je analiza podataka, a izvještaji su kreirani pomoću alata iReport. Na primjeru korištenja tih alata objašnjene su značajke alata otvorenog koda za poslovnu inteligenciju.
|
Z. Milokanović (Vipnet d.o.o., Zagreb, Croatia)
Izvedba skladišta podataka za uslugu kratkih poruka
Sažetak – Telekomunikacijske operatere karakterizira kompleksnost i mnogobrojni sustavi koji generiraju veliku količinu podataka o aktivnostima korisnika. Kako bi mogli iskoristi znanje o korisniku i pravovremeno reagirati na tržišne izazove operateri uvode koncept skladišta podataka kao temelj za potporu poslovnog odlučivanja. Ovaj rad prikazuje izvedbu skladišta podataka za uslugu kratkih poruka kao jedne od najpopularnijih usluga u telekomunikacijama. Opisan je osnovni dimenzijski model čije se tablice pune produkcijskim zapisima. Kako takav model u velikom broju slučajeva ne pruža optimalne performanse prikazana je najčešće korištena metode optimizacije kojom se postižu brži upiti za dohvat željenih podataka.
|
|
Basic information:
Chairs:
Mirta Baranović (Croatia), Matteo Golfarelli (Italy), Boris Vrdoljak (Croatia), Roberto Sandri (Croatia)
Program Committe:
Alberto Abello Gamazo (Spain), Marko Banek (Croatia), Mirta Baranović (Croatia), Ladjel Bellatreche (France), Alfredo Cuzzocrea (Italy), Todd Eavis (Canada), Dragan Gamberger (Croatia), Matteo Golfarelli (Italy), A. Min Tjoa (Austria), Jesus Pardillo (Spain), Zoran Skočir (Croatia), Mladen Varga (Croatia), Boris Vrdoljak (Croatia), Robert Wrembel (Poland)
Scope:
The term Business intelligence (BI) refers to applications and technologies used to gather, provide access to, and analyze data and information about company operations in order to help them to have a more comprehensive knowledge of the factors affecting their business. BI systems usually lean on data warehouses (DWs) to guarantee an easy and quick access to information. The widespread deployment of these systems is making users more and more demanding in terms of applications, types of analysis and performances; on the other hand ten years of research in the data warehousing field paved the way for new applications that goes beyond traditional OLAP analysis and concern real-time, spatiotemporal as well as distributed data. Furthermore, BI research is now dealing with the integration of BI with Corporate Performance Management (CPM), balanced scorecards, and Business Process Management (BPM).
Like the previous successful editions, miproBIS 2011 provides an international forum where both researchers and practitioners can share their findings and practical experiences in the field of BI. In particular, besides submissions in traditional DW and OLAP topics, we encourage papers concerning emerging research topics and applications in the DW, OLAP and BI area.
The scope of the conference includes, but is not limited to:
- Data warehouse design and schema evolution
- Multidimensional modelling and queries
- Data warehousing consistency and quality
- Online Analytical Processing (OLAP)
- Tools for data warehousing and OLAP
- Integration of data warehousing and data mining
- Metadata management
- Integration of Business Process Management (BPM) and Business Intelligence (BI)
- Forecasting and simulation in Business Intelligence
- Corporate Performance Management (CPM) and Business Intelligence
- Balanced scorecards
- Decision Support Systems
- Active/real-time data warehousing
- Cloud intelligence
- Data warehousing with unstructured and semi-structured data (e.g. XML)
- Data warehousing and industry applications (e.g. ERP, CRM)
- Data warehousing in mobile/wireless environments
- Biomedical data warehousing
- Sentiment analysis
- Analytics for semantic Web
Official languages are English and Croatian.
Location:
Opatija, often called the Nice of the Adriatic, is one of the most popular tourist resorts in Croatia and a place with the longest tourist tradition on the eastern part of Adriatic coast. Opatija is so attractive that at the end of the 19th and beginning of the 20th centuries it was visited by the most prominent personalities: Giacomo Puccini, Pietro Mascagni, A. P. Čehov, James Joyce, Isidora Duncan, Beniamino Gigli, Primo Carnera, Emperor Franz Joseph, German Emperor Wilhelm II, Swedish Royal Couple Oscar and Sophia, King George of Greece.
The offer includes 20-odd hotels, a large number of catering establishments, sports and recreational facilities.
For more details please look at www.opatija.hr/ and www.opatija-tourism.hr/.
Registration / Fees:
|
REGISTRATION / FEES
|
Price in EUR
|
|
Before May 09, 2011
|
After May 09, 2011
|
| Members of MIPRO and IEEE |
180
|
200
|
| Students (undergraduate), primary and secondary school teachers |
100
|
110
|
| Others |
200
|
220
|
Contact:
Boris Vrdoljak
Faculty of Electrical Engineering and Computing
Unska 3
HR-10000 Zagreb, Croatia
Phone: +385 1 6129 756
Fax: +385 1 6129 915
E-mail: boris.vrdoljak@fer.hr
|
|
|
|
| Currently there are no news |
|
|
|
|
|