Libcolumbus is a small error tolerant search engine designed to deal with noisy data and typos. It will power the searches in the next generation of Ubuntu's HUD system as well as other searches. It has a fast implementation of the Levenshtein distance algorithm, which allows it to correct errors such as added and dropped letters (e.g. 'bar' -> 'bard'), changed letters ('ctr' -> 'car') and translations ('acr' -> 'car'). It also allows the user to customize the error values. Libcolumbus is designed to be small, efficient and easy to embed. It is programmed in C++ but also provides C and Python APIs.
|Tags||search engine Error correcting|