Automated Quality Assurance for Heuristic-Based XML Creation Systems

Large volumes of XML document instances are created by conversion systems that rely on heuristic-based tagging. Quality assurance is typically conducted on individual document instances, but often there is no systematic regression testing of the conversion system applications used to tag the instances, especially when the patterns used for heuristic-based tagging are changed or updated. Because XML conversion systems have a deterministic output for a given input document, automated quality assurance can be used to ensure the stability of markup creation by a conversion application over a large volume of documents in a conversion system. The automated test suite can also be used to validate general improvements to processing heuristics when introducing broad-based changes to the overall system This paper examines the issues that are encountered when developing conversion systems, especially those systems that rely on heuristic-based tagging, and the quality problems that arise when upd...

• Type: Conference
• Year: 2004
• Where: EXTREME
• Authors: Bruce Rosenblum, Irina Golfman
• Keywords: Conversion Systems, EXTREME 2004, Heuristic-based Tagging, Natural Language Processing, Quality Assurance