Efficient schema extraction from large XML documents

Although the presence of a schema enables many optimizations for operations on XML documents, several studies have shown that many XML documents in practice either do not refer to a schema, or refer to a syntactically incorrect one. It is therefore of utmost importance to provide tools and techniques that can automatically generate XML Schema Definitions from sets of sample documents. While previous work in this area has mostly focused on the method based on regular expressions, we consider its many inadequacies. We provide a theoretically complete algorithm that always infers the correct XSDs when a sufficiently large corpus of XML documents is available. In addition, XTree impressively minimizes the necessary time and main memory to extract the schema. Our approach features several advantages over known techniques: XTree scales to very large documents (beyond 1 GB) both in time and memory consumption; it is able to extract a general, complete, correct, minimal, and readable schema for complex documents; it detects elements appear as a sequence or choice. Experiments confirm these features and properties.