Comparing of SGML documents

Documents can be represented as structures with a hierarchial arrangement of text and non-text nodes, where nodes are labeled by category names such as paragraph and section. Representing documents this way is a natural consequence of using the Standard Generalized Markup Language(SGML) to encode text documents which has many applications in different areas. There are many circumstances in which one structured document is a variant of another and we need to compare them to find the relationships (editing changes) between them. One simple way to do that is to model SGML documents by ordered labeled trees and then find a mapping between them. In this proposal we consider this problem deeply, and mention some solutions from the literature. Also, we introduce a research area which consists of a better approach for modeling the SGML documents and related problems. Finally, we mention some methods for attacking to these problems. keywords: SGML, Comparing of documents, Ordered Labeled Trees, Tree-to-tree correction, Directed Labeled Graphs. 1 Motivation and Significance of the problem Documents can be represented as structures with a hierarchial arrangement of text and non-text nodes, where nodes are labeled by category names such as paragraph and section. Representing documents this way is a natural consequence of using the Standard Generalized Markup Language(SGML) to encode text documents. SGML or other variants such as HTML are widely used and even documents that are not simple hierarchies can be represented using SGML. One simple way to model the documents represented in SGML or HTML are trees with labeled nodes where the left to right ordering of the offspring of a node is significant. We call this tree an ordered labeled tree. However, there are other sophisticated ways to model SGML documents which we consider later. There are many circumstances in which one structured document is a variant of another. For example, there may be several manuscript or printed versions of an existing text; there may be several translations of a text; there may be several machine-readable forms of a text produced for different audiences(e.g. a text with minimal apparatus for students and a more extensive apparatus for researchers) or finally there may be interest in maintaining several versions of a document that is being produced in a cooperative manner by several authors. In all of these circumstances, we need to compare different variants of a document and find the relationships (editing changes) among them and in most of the cases these variants have a hypertext structure such as SGML. Finding the editing changes also can be used is other applications like: