A Method of Recommendation the Most Used XML Tags

Abstract

Processing of a large data set which is known for today as big data processing is still a problem that has not yet a well-defined solution. The data can be both structured and unstructured. For the structured part, eXtensible Markup Language (XML) is a major tool that freely allows document owners to describe and organize their data using their markup tags. One major problem, however, behind this freedom lies in the big data retrieving process. The same or similar information that are described using the different tags or different structures may not be retrieved if the query statements contains different keywords to the one used in the markup tags. The best way to solve this problem is to specify a standard set of the markup tags for each problem domain. The creation of such a standard set if done manually requires a lot of hard work and is a time consuming process. In addition, it may be hard to define terms that are acceptable by all people. This research proposes a model for a new technique, XML Tag Recommendation (XTR) that aims to solve this problem. This technique applies the idea of Case Base Reasoning (CBR) by collecting the most used tags in each domain as a case. These tags come from the collection of related words in WordNet. The WordCount that is the web site to find the frequency of words is applied to choose the most used one. The input (problem) to the XTR system is an XML document contains the tags specified by the document owners. The solution is a set of the recommended tags, which is the most used tags, for the problem domain of the document. Document owners have a freedom to change or not change the tags in their documents and can provide feedback to the XTR system.