Weekly Report (Abdullah Cemil Akçam/Nov 18 - Nov 25)

this week i have mainly focused on java platform. i have made some research on the net for java and java related technologies. in today's software world, integrated development environments (ide) are vital for good projects. for an ide to work properly, needed plugins, packages and extensions should be installed beforehand. for java, the most famous ide is netbeans which is a product developed by Sun. Java platform and java programming language are also developed by Sun microsystems. in short, one can easily find java related technologies developed by Sun for all purposes. there are some third party products but in today's world integrity and synchronization are key concepts for good projects. as a result, products developed by a single company is so important. more detailed information can be gathered from :

http://www.sun.com/software/


for the ide part i have preferred netbeans. for netbeans to work properly, Java Development Kit (jdk) and java runtime environment (jre) packages must have been installed previously. thereafter netbeans is installed. to develop web application in netbeans an application server is needed. actually, application server is needed for all platforms in today's dynamic web application. for example in visual studio sql server is used for this purpose. for java there are many alternatives for application server part. some popular candidates are tomcat, jboss. i have chosen glassfish v3 for this part. it is also a sun product. more information about it can be found at:

https://glassfish.dev.java.net/downloads/v3-techPreview-1.html

this installation period is not painless but once you are done with all the parts successfully, the remaining part is just fun with the nice java style and user-friendly tools


Abdullah Cemil Akçam

Weekly Report (Nov. 18 - Nov. 25 / Cihan Bahran)

This week I made further research about data mining and read more about cluster analysis and classification methods.
I also checked the slides of the previously given graduate data mining course in our department, Ceng 714. Slides provide a quicker understanding of the basic points. I also downloaded slides of similar courses from some universities like Stanford, MIT, University of Illinois Urbana Champaign and had a clearer idea about what lies at the heart of data mining & machine learning algorithms and their current widely used applications.
I think my research thus far will prove very useful as a guide for our prototype design in this semester and the final design in the second semester, though they constitute only the tip of a huge iceberg.

Weekly Report(Nov 11-Nov 18/A.E.Sarıyüce)

• Hi , next week comes and the new weekly reportis here again.. In this week, I continued to study basics of data mining , attended a meeting with the Azerhan Özen, co-founder of Butun-yd Erp Consultancy co. (http://www.butun-yd.com/index.html) and wrote some parts of Requirements Analysis which we submitted yesterday. About data mining, I focused on BI aspects of algorihms and methods in data mining rather than pure data mining subjects. Since we do not need to reinvent the wheel, it is better to study on predetermined specific BI-oriented data mining algorithms. Last friaday, I went to the office of Butun-yd Erp Consultancy co. at Kızılay and meet with Mrs. Özen for 2 hours. I were not thinking the meeting will be so useful for me, however it was very benficial for me to see a real implementation of our project in the market. For 2 hours, Mrs.Özen talked me about their ERP system, named Canias-ERP, and show the several features of this system on computer. Their system just takes the raw data and outputs very meaningful information. No data mining, no complicated algorithms, even no serious database model. Just bunch of graphs, visual charts and visual aids whick makes the job of bosses easier. Lastly, we wrote the Requirements Analysis of our project and submitted it yesterday. I wrote the functional requirements part and draw many data flow diagrams. I think that further we proceed in the project, ideas become more concrete and implementable.. See you next week.. Ahmet Erdem SARIYÜCE

Weekly Report (Hakkı Büyükcengiz/Nov 11 - Nov 18)

Hi all,

This week i thought and researched about some technical aspects. I reseached sun technologies, such as java, j2ee, JSF, mysql, netbeans and glassfish. I concluded that: all sun technologies "highly compatible" with each other, and as other advantages, they are open source, free and platform independent. When an SME is considered, these features of sun technologies become more appealing, because an SME may not cope with the cost of other commercial products or other commercial platforms.

Also, this week i researched some open source and free data mining tools from the web. I've found Java Data Mining API, this api seems to be very advanced for us now, but it is worth to read it from here : http://www.artima.com/lejava/articles/data_mining.html at least, i think that it will bring us new ideas about this issue.

-Hakkı Büyükcengiz.

weekly report (Nov. 11 - Nov. 18)

This week I have made some research about how big companies proceed in their BI projects. Their main approach is dividing the process into modules and for each module preparing a production. This is both for technical ease and for financial gain thats they can demand money for each product. When I try to project this approach into our design project, I came up with a modular process model. I call it "process framework". I proposed this idea in our project group and other group members accepted it. So we will proceed according to 4 phased process framework. The details are in srs document.

Other than this research, I went to "makim" which is the substitute sme for our project. I have spoken to the owners of the company tried to understand their needs. I also tried to predict what kind of BI product works properly in this kind of corporations. Which process model is best was another question that I tried to answer. My conclusion can be seen in srs document.

Last of all, me and other friends have worked on srs document whole week. Each group member have completed his part.


Abdullah Cemil Akçam

Weekly Report(4 Nov. - 11 Nov. Cihan Bahran)

This week I was mostly involved with doing literature survey about the requirements analysis report. I analyzed the existing business intelligence products for SMEs on the market and compared them to our project.

Moreover I started reading the "Machine Learning and Data Mining" book by Igor Kononenko and Matjaz Kukar. It briefly introduces the data mining methods and I think Machine Learning is an interesting subject and may find its use in our project.

Weekly Report (Hakkı Büyükcengiz/Nov 4 - Nov 11)

Hi all,

This week i couldnt have much time to research, but i glanced at what data mining is and is not, the relation between statistics and data mining i.e. the differences and similarities. I researched these issues from books and mostly from WEB. I learned more "concrete" techniques and cases about data mining, but most of these techniques for retailing sector i.e. on CRM and customer profiling , so if our SME is not a retailing company, i think there are other aspects to be thought of. Now, i research on this, for an SME producing intermediate products, which data patterns could be used more, and which results could be desired to be seen by the user.

-Hakkı Büyükcengiz.

weekly report - (Nov. 4 - Nov. 11 )

in this week I did not do much thing for the design project. I have made a telephone communication with the factory manager and waiting for the reply right now. Moreover, I have studied on last year's srs samples from senior.ceng.metu.edu.tr. I have studied on our srs document and tried to determine the requirements.


Abdullah Cemil Akçam

Weekly Report (Hakkı Büyükcengiz/Oct 28 - Nov 4)

I research mining and BI so i took related books from METU library, and now i am still reading two books named "BI for Customer Relationship Management" and "Knowledge Management and Business Intelligence". Also, we talked with the manager of SME, this was the first meeting so we asked what kind of a result they desire to see as a solution, roughly. Of course, there is always data to mine somewhere around, so i searched for "valuable" information sources from the web, one of these is www.tuik.gov.tr . On this web site, i see that huge amount of information is available, so relevant information can be used to mine as well, for example, all related producta selling information, and even purchasing power of these products of Turkish people in last 6 months! Furthermore, we had a talk with Erol Sayın from Industrial Engineering, he gave us a lot of valuable information. Also, i studied and learned some economical terms and concepts in order to find and analyse more information from the WEB, because our project have lots of economical aspects so i think that economical rules are very important to come up with a useful product.

-Hakkı Büyükcengiz.

Weekly Report (Oct 28 - Nov 4)

In this week, I have made research about the BI products on current market and read several tutorials about data mining. I've looked for IBM Cognos, SAP and Oracle to inspire from them in our project. Actually, when we are searching for an advisor company, some of them said that they already use some-kind of a system we plan to do. SAP is prevalent in middle-sized companies whereas Oracle is common at the much bigger companies. I think learning the algorithms and design choices made in these products can give inspiration to our project as well. On the other hand, I started to study data mining in this week. I think data mining is very important application area in computer science rather than a pure research topic. Extracting meaningful and useful patterns from large data sets explains the main idea of data mining. In business intelligence, obtained meaningful patterns from huge data provides people to estimate about future plans and it explains why the BI is vital for enterprises. After reading some material about data mining, I realized that algorithms we are to use on data mining phase of project will constitute the skeleton of whole work. As we talked in meeting, we will share the subjects about data mining to proceed more efficiently. In the next week, I plan to continue studying data mining . Also we will meet with our advisor company about SRS (Software Requirements Specification).

Ahmet Erdem SARIYÜCE

weekly report - (Oct. 28 - Nov. 4 )

This week I have read the book "Business Intelligence Roadmap" . In this book, requirements of a business intelligence solution process are explained in a step by step manner. I have studied about these steps. Moreover, me and Hakkı have gone to Tuğcular and met with the factory manager. Our meeting was efficient, the manager's attitude towards us was very good. He promised to do best for our project but he said that he has to debate about this project with the owners of the factory. He wanted us to send him an e - mail including the data that we will demand from them. His hesitation was about the private data. I send the mail and now waiting for the reply.

Before sending the mail, I have searched for valuable data for our project and I have concluded that with any information collected, we can make good queries thanks to data mining. For example, if the factory owners do not want to show their monthly budget, then we can present them some analysis about possible market shares. In my opinion, if we can provide a running demo which gives some hints about the whole project, then the owners will demand some queries from our program and the project will hit the target. Now I am thinking about how to realize this running demo.


Abdullah Cemil Akçam