On Hidden Disorder in XML Tags

Automatic integration of structured and semi-structured data has been a goal of nearly thirty years of integration research. The eXtensible Markup Language (XML) integration attempts have not succeeded either. The lack of progress demonstrates early assumptions were naïve. This paper will show XML schemas possess natural language characteristics that pose significant barriers to achieving full automatic integration of semi-structured data. This aspect of XML evolution has not been addressed in the literature. XML exhibits many of the same ambiguities as natural language. Automating XML integration puts forward a challenge similar in magnitude to automated natural language translation efforts. Correct and complete integration needs to satisfy meaning-preservation constraints; e.g., a mapping function that is invertible, proof preserving, vocabulary preserving and structure preserving. Natural language characteristics in XML can theoretically be utilized to predict and explain the level of XML automated integration possible.