What is most valuable?
Weka is a very easy to use Data Mining solution. It is great for learning and for doing small experiments before exploring the data deeper. Another important feature is the number and diversity of algorithms that make Weka an excellent solution for rapid testing. The several interfaces that are provided also allow a diverse range of applications and uses. Another important aspect is the ability to easily integrate new algorithms to the solution, and also it’s integration in terms of Java code.
How has it helped my organization?
I have used Weka both in teaching and in industry projects, for several types of Data Mining tasks.
What needs improvement?
Scalability and performance are the main aspect of improvement in Weka, since it has the main Java limitations, regarding the JVM. Besides that, the pre-processing part of Weka is the hardest to use aspect of it.
For how long have I used the solution?
I've used it for more than 10 years.
What was my experience with deployment of the solution?
No issues with deployment.
What do I think about the stability of the solution?
No issues with stability.
What do I think about the scalability of the solution?
Yes, fine tuning the JVM memory is something to be careful about.
How are customer service and technical support?
It is open source and customer service is not something that is given. However, there is a community of Weka users and some documentation that can help with the use of Weka. Technical Support
Same thing as the customer service.
Which solution did I use previously and why did I switch?
I used the old Clementine solution (now in the IBM portfolio). Weka ends up being more versatile, both in terms of diversity of algorithms, integration flexibility and there are less costs.
How was the initial setup?
The setup is straightforward, just download and start using.
What about the implementation team?
I implemented in-house and for other companies.
What was our ROI?
High, since it has low costs and is very easy to use.
Which other solutions did I evaluate?
Yes, but long ago. I evaluated Oracle Data Mining, Clementine, and SAS Enterprise Miner.
What other advice do I have?
Data Mining know how is needed to use the solution, but that is what is expected, since this tool is for Data Scientists.