Attribute Based Ranking Search Engine Project

Our ranking engine has reached a major mile stone. It is now finally a reality and fully functional (least it was until they removed the NPAPI interface). This is due in great part to the help I have received on stackOverflow; particularly from BalusC. I just wanted to thank everyone for their help; direct or indirectly through answering other people's questions. This project would never have reached this point were it not for your help! It is truly greatly appreciated.

There is still some work to finish off the web site, but the actual engine side, and its web pages are now fully functional. The engine is currently ranking 1225 digital cameras using 89 attributes.

About us

We are a small garage project of two; myself and my partner. My partner has provided the needed investment money for the Australian provisional patent and has supported me while I developed the project for the last six years. I am the only developer and have done all of the coding myself.

How did the engine come about?

I have had permanent earaches for over 14 years. Rather than sit around and feel sorry for myself, I decided to try and implement a Ranking Engine design that I first envisioned, over eight years ago, and which is based on an abstract theory for how the human brain uses neural networks to rank things for us.

Given that I was an applications programmer and knew very little about web development, the steep learning curve for JSF has been considerably eased by stackoverflow's collective knowledge.

About the ranking engine

The engine can accurately capture ranking criteria for virtually anything the human brain can rank and then just as accurately emulate the brain's ranking processes using that criteria. Like the human brain, the engine can rank things that have no attributes in common within the ranking criteria. It can also use attributes for which we have no formal measurement and which the brain uses on a daily bases. This is possible, because of the use of graphs to store and edit the criteria.

The software's most powerful feature is that it can capture a knowledgeable expert's ranking criteria knowledge and make it available for anyone to use at the click of a button.

The project

According to the WordCount plugin for Netbeans (excluding Netbeans generated code): the Java editor to create the initial ranking files is 77,710 lines of code; the Java applet to allow users to customise the ranking criteria is 64,839 lines. The web site is now 36,459 lines of JSF 2.0 and HTML code. The backing bean and servlet code total 34,143 lines.

Thanks again to all those who have helped to get us this far, and thanks in advance for what is yet to come. No doubt I will be back with more problems.

Kind Regards, Robbie