Lexical Error Compensation in Handwritten-Based Mathematical Information Retrieval
Seyed Ali Ahmadi, Abdou S. Youssef · Czech digital mathematics library · 2008
Abstract. Entering mathematical queries, in general, can be a demanding task. Mathematical notation is two-dimensional and cannot be easily typed with a standard QWERTY keyboard. Handwriting appears to be the most intuitive and promising method to express mathematical queries. Recognition technology for handwritten mathematical notation has never been applied in math search. The objective of this research is to design and implement an automated symbol-recognition error compensation system for the handwritten-based query processing. To achieve this objective, we have designed and implemented a query processing algorithm which modifies and expands handwritten queries using different policies, in order to compensate for lexical symbol-recognition errors. Experiments have demonstrated that these policies substantially improve the performance of a handwritten-based math IR system. Key words: mathematical information retrieval, handwritten mathematical recognition, digital libraries, math search, error compensation 1