This paper presents a new version of an evolutionary algorithm that cre-ates XSLT programs from scratch, given a single example of their intended input andoutput. XSLT is a general purpose, document-oriented functional language, generallyused to transform XML documents (or, in general, solve any problem that can be codedas an XML document). Previously, an efficient solution to the problem was proposed; inthis paper, we improve on those results by testing different fitness functions, adding anew operator and changing the type of output document that can be obtained. Resultsshow that the best results are obtained without considering the length of the XSLTprogram and including this new operator. The new type of output files used is alsomore realistic, and improves on old results.
展开▼
