Language identifier demo

Enter a Latin-transliterated text (e.g. English, French, Ugaritic, Phoenician) to find its closest language profile based on character frequency ordering.

To go back to the previous page, click here.

Hi, this text is in English but you can paste anything you want here instead. The point of this demo is to show how character frequency will detect the language of the text you will have inputted. It works simply by counting every distinct character and listing them in descending order of frequency (DFO). It then compares that list of characters to the precomputed DFOs that the program already has in memory for various corpuses or inscriptions written in different languages. It then returns a list of the profiles that match best using Spearman's rho with 1 being 100% match and 0 being no match. Make sure your text isn't too short as this technique works best for texts of at least 150-300 characters or more. If you are lucky, the program has your language in memory and will list it somewhere at the top of it's prediction! This program runs entirely in your browser and does not require an internet connection after loading. Analyze Text