Context-Sensitive Spelling Error Correction Using Inter-Word Semantic Relation Analysis

Error words that appear in Korean texts can be largely categorized into non-word spelling errors and context-sensitive spelling errors. Of the two, context-sensitive spelling errors are shown only when considering the meaning of the word in the given context and its syntactic relation, and they are the most difficult to correct among spelling errors. Context-sensitive spelling errors can be categorized into homophone errors, typographical errors, grammatical errors, and cross-word boundary errors. To correct context-sensitive spelling errors that occur due to typographical errors, this study proposes a statistical context-sensitive spelling check using inter-word semantic relation analysis. With confusion sets created in advance, we can find and correct context-sensitive spelling errors using reliability based on the conditional probability and chi-square statistics between each word of the confusion sets and the context as well as the typing error rate. As a result of applying the proposed method, all 5 confusion sets showed higher precision (92.68%) and recall (83.95%) than the baseline (precision 80%, recall 80%).