Spell-checking queries by combining Levenshtein and Stoilos distances
Zied Moalla, Lina Fatima Soualmia, Élise Prieur-Gaston, Thierry Lecroq, Stéfan Jacques Darmoni · F1000Research · 2011
We propose in this paper a simple yet efficient method in order to correct misspellings of queries submitted by users to an online search tool in medicine. In addition to exact phonetic term matching, we test two approximate string comparators: the string distance metric of Stoilos and the Levenshtein edit distance. We propose here to combine them. At a threshold comparator score of 0.2, the normalized Levenshtein algorithm gives the highest recall of 76% but the highest precision 94% is obtained by combining the two distances of Levenshtein and Stoilos. Despite the well-known good performance of the normalized edit distance of Levenshtein, we show in this paper that its combination with the Stoilos algorithm improves the results for misspelling correction of user queries. This method may be applied to text documents in Electronic Health Records or clinical documents.