As XML (extensible Mark-up Language) is gaining its popularity in data exchange over the Web, querying XML data has become an important issue to be addressed. In native XML databases (NXD), XML documents are usually modeled as trees and XML queries are typically specified in path expression. The primitive structural relationships are Parent-Child (P-C), Ancestor-Descendant (A-D), sibling and ordered query. Thus, a suitable and compact labeling scheme is crucial to identify these relationships and henceforth to process the query efficiently. The authors propose a novel labeling scheme consisting of to support all these relationships efficiently. Besides, the authors adopt the decomposition-matching-merging approach for structural query processing and propose a hybrid query optimization technique, Twig INLAB to process and optimize the twig query evaluation. Experimental results indicate that Twig INLAB can process all types of XML queries 15% better than the Twig Stack algorithm in terms of execution time in most test cases.