Error detection, error correction and performance evaluation in on-line mathematical expression recognition

Automatic recognition of on-line mathematical expressions is difficult espe cially when there exist errors. In this paper, we incorporate an error dete ction and correction mechanism into a parser developed previously by us bas ed on definite clause grammar (DCG). The resulting system can handle lexica l, syntactic and some semantic errors. The recognition speed for 600 common ly seen expressions is quite acceptable, ranging from 0.73 to 6s per expres sion on a modest workstation. In addition, we propose a performance evaluat ion scheme which can be used to demonstrate the effectiveness of both the s ymbol recognition and structural analysis stages by a single measure. (C) 2 001 Pattern Recognition Society. Published by Elsevier Science Ltd. Al righ ts reserved.