FW Editor: What are your plans or objectives in the near future?
Zoran Sevarac: Our objective is to create rich Java neural network IDE and research platform. We want to provide tool that is easy to use for Java developers and which will allow them to drag n drop neural networks in their apps. For advanced users and researchers in the field of neural networks, we want to provide easy way to customize and experiment with different types of neural networks. Also we plan to provide more samples and interactive tools which will make Neuroph an educational platform for neural networks.

FW Editor: How and when did you start writing the code for Neuroph? What inspired you the most? Do you plan to develop new software, or are you more focused on optimizing the current ones?
Zoran Sevarac: It all started as a working assignement at my student days when I got interested in neural networks about 5 years ago. I started working actively on it since end of 2008 when I published it on SourceForge. Many contributors helped to make it as it is now. From the beginning, I was inspired by the idea of making the visual component based artificial brain builder that allows to experiment with various ideas related to artificial intelligence. It was kind of very cool idea :) Along the way we also developed some very practical tools like image recognition, OCR, stock prediction etc. In future I'll be focused on adding new features and optimizing this software, and we have many plans in that direction.

FW Editor: Do you plan to improve or change or improve Neuroph in any way? If yes, how are you going to do that?
Zoran Sevarac: Of course, we're always improving something :) And there are endless posibilities. Two main things that are ongoing development are: 1. Porting Neuroph GUI to NetBeans platform which will provide us state of the art visual GUI http://sites.google.com/site/nbugserbia/projects/porting-neuroph-to-netbeans-platform 2. Integration with Encog project and using Encog high performance core, which will provide great performance improvements. http://netbeans.dzone.com/encog-neuroph-collaboration I expect those will be published till the end of the year.

FW Editor: Neuroph has quite a strong competition but, so far, it succeeded in making a name for itself. What do you think about the huge success of Neuroph?
Zoran Sevarac: Neuroph has quite a strong competition but, so far, it succeeded in making a name for itself. What do you think about the huge success of Neuroph? I think its really great thing, and I must I admit its more then I expected in beginning. On the other hand I think it succeeded thanks to its quality and good project management. I'm listening to our users and developers, I really want to hear what they think and apply those suggestions. Also the project had the clearly defined goals and vision from the beginning, and basicly those are the same today.

FW Editor: What are the main advantages of using Neuroph over any other similar product?
Zoran Sevarac: Easy of use, nice friendly GUI and API, good documentation and support.

FW Editor: Besides Neuroph, is there any other similar application that had a positive impact over you? Why?
Zoran Sevarac: Well those are not similar but I would like to mention: JUNG, great Java graph library, very easy to use NetBeans, IDE and Swing app framework, saves development time, gives proffessional look and architecture to our apps jMonkeyEngine, Java 3D Game Engine, unbeilivable what can be made in java Robocode, small tank battle simulation, very cool idea!

FW Editor: What is your favorite Neuroph feature and why?
Zoran Sevarac: Well, at the moment it is image recognition support. Although it is very basic and simple, that was the first real world application of Neuroph and after that things started to happen. Before that it was all kind of academic excersize, but when I saw that someone really used Neuroph and contributed to the project, I thought to myself: hey this is actually working :) It also evolved into some basic OCR support, and we'll improve both in future releases, since those two along with prediciton, are the most wanted application for neural networks.