Logic compression of dictionaries for multilingual spelling checkers

Boubaker Meddeb Hamrouni · 1994

To provide practical spelling checkers on micro-computers, good compression algorithms are essential. Current techniques used to compress lexicons for indo-European languages provide efficient spelling checker. Applying the same methods to languages which have a different morphological system (Arabic, Turkish,...) gives insufficient results. To get better results, we apply other "logical" compression mechanisms based on the structure of the language itself. Experiments with multilingual dictionaries show a significant reduction rate attributable to our logic compression alone and even better results when using our method in conjunction with existing methods.

Read the paper · More papers on PaperTik