Software

Scientio, Inc announce demo versions of two new products, XML Miner and XML Rule, that together offer a COM component solution to data mining XML data, are available for download from the site http://www.xmlminer.com/downloads.asp.

XML Miner analyzes and data-mines data in XML Format, thus avoiding the constraints of proprietary data formats, but also extending data mining away from relational representations to hierarchical, tree based structures. Using Fuzzy Rule induction technology, XML Miner generates a rule set that explains and predicts selected data items in an input data set, defined using XPath expressions, based on other user selected items. The resulting rule set is expressed in Metarule, a dialect of XML that can be easily converted to English if.. then type rules using an XSL transform supplied, and is also understood by the runtime processor XML Rule. Using this runtime component users can embed a solution anywhere that takes a COM object. A white paper detailing the above is available at: http://www.xmlrule.com/XML%20Miner%20White%20Paper.pdf

Various online demonstrations of XML Rule and XML Miner are available at the Scientio web site including a recent study on predicting the value of works of art at auction, that demonstrates XML Miner and XML Rule performing comparably with professional art valuers in accuracy. http://www.xmlminer.com/online.asp

With XML now becoming the standard for transferring information between disparate applications and computers across the Internet, XML Miner and XML Rule have application in business to business transactions and adding intelligence to web sites for customer classification, pre-filtering and behavior prediction.

XML Miner can handle categorical and numerical data types for both input and output and can auto-discover the type if required. The Metarule format can represent a large range of knowledge paradigms, including neural nets and Markov models. http://www.xmlminer.com/whymetarule.asp

XML Miner and XML Rule are the first of a planned range of software components encompassing various knowledge extraction techniques based around Metarule and XML.

Copious examples are supplied on the web site and with the demonstration download, illustrating the use of these components with Java, C++ and VB, on web servers, web browsers and in conventional applications. http://www.xmlminer.com/resources.asp

Scientio Inc, with offices in Belle Harbor, NY and Buckinghamshire, UK, specializes in innovative component software making use of Computational and Artificial Intelligence. Web: http://www.xmlminer.com Email: info@metadatamining.com