Language Guesser finds language markers in the text thus initiating the process of accumulating the weight of languages. Each language sums up marker weight multiplied by number of marker occurrences in the text. Not all the words are involved into the identification process, for instance, the algorithm eliminates prepositions, conjunctions and words shorter than four characters. When the algorithm work is completed, the Language Guesser compares weight values and shows the table of languages with a probability index.
The identification process can introduce some errors that depend on language specific features, uniqueness and number of words in downloaded text. Therefore, the Language Guesser results in a list of the most appropriate languages identified for downloaded text.
