Vijay SankaranDirector, Application Development| Ford Motor CompanyAnn Arbor, Mi, United States
Everyone is always looking for new and improved BI tools. Share your experiences on which tools you think are useful, which ones suck, and common problems you've had using tools... Saving Changes...
We're (Teradyne) currently going through a BI/Portal tool selection and have evaluated many different vendors at this point. For the BI side of the project we've narrowed it down to Microstrategy, Business Objects, Brio and Oracle.
Any stories/experiences with these companies would be appreciated :)
I have had the opportunity to work with the Cognos (www.cognos.com) Impromptu and PowerPlay tools and found them to be well suited for end-user ad-hod reporting and OLAP respectively.
I have just started working with Oracle Discoverer to access Oracle Applications data through NoetixViews and find it to be less user friendly and less powerful than Impromptu. We are using it because of the Oracle Applications implementation we have underway. At first, it seemed liked a good idea. On the surface, the Cognos tools and the Oracle tools appeared similar and with Oracle, its a single-vendor solution. But now, I would almost rather have Impromptu and PowerPlay.
Anyone out there with Noetix advice to share? Saving Changes...
Anonymous
Dear Friends, I am a Software professional with the Skillsets of Java and ASP, i need your replies, on the tools to be used for the appropiate projects. At present I hold the post of Project Co-Ordinator in Paramount Technologies, Chennai and leading a team of nine members,. The project is in ASP and SQL Server7.0.
Thank you Guys & Gals.
- Sabari Saving Changes...
Anonymous
Dear Friends, I am a Software professional with the Skillsets of Java and ASP, i need your replies, on the tools to be used for the appropiate projects. At present I hold the post of Project Co-Ordinator in Paramount Technologies, Chennai and leading a team of nine members,. The project is in ASP and SQL Server7.0.
Dear Friends, I am a Software professional with the Skillsets of Java and ASP, i need your replies, on the tools to be used for the appropiate projects. At present I hold the post of Project Co-Ordinator in Paramount Technologies, Chennai and leading a team of nine members,. The project is in ASP and SQL Server7.0.
Folks, We evaluated business object, cognos and micro strategy. We are using cognos and are converting some users over from Business objects. One importnat decision to be taken up front is how heavy you want the client to be. We are going with very thin clients and are moving most users over to Web access. Only a small number will have heavy clients. Of course the sys admins, cube designers etc will all use developer seats with full client access. On the Data Mining side I would recommend steping out of the suiet tools to stronger products, currently I use Intelegent Data Miner and SAS. Please note these are my personal opinion and are given without commitment or prejudice. Saving Changes...
Vijay SankaranDirector, Application Development| Ford Motor CompanyAnn Arbor, Mi, United States
One of the biggest problems with data mining tools that I've seen is the lack of integration with the RDBMS. Most times you have to extract out into a flat file and then use a data mining product. Does anyone know any good embedded data mining products that runs directly against RDBMS? There are some that use Decision Trees directly against SQL databases but very few that use nearest neighbor or neural nets against a relational RDBMS. Saving Changes...
Michael WoodProject Manager / Business Analyst / Business Process Improvement Guru| Independent ContractorGig Harbor, Wa, United States
There is a good reason why you do not find mainstream data mining and data warehouse products that use an RDBMS. Most RDBMS systems are geared for transaction processing. This means fewer indexes and normalized data sets (tables, files, etc.). Data Warehouse and Data Mining systems rely on lots indexes and denormalized structures. They are built to navigate huge quantities of data quickly against any number of where clauses. For example: A large customer data base took 151 hours per complex query using a normalized SQL structure. That same data base in a denormalized structure could do virtually any search against 1.5 million customers in under 1 minute. This is a huge area of debate amongst DBA's and in the industry in general. If you are using an RDBMS then I suggest you imbed the views into the tables and index the heck out of them. RDBMS products that support inverted key lists work best for data warehouse and mining applications. The best advice is to stay out of the theoretical and stick with the physics of how data is physically managed on a disk. Spread the data across as many disk spindles and controllers as possible. And stick with what works. Saving Changes...
"Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining."
