{"id":171345,"date":"2014-04-22T10:07:06","date_gmt":"2014-04-22T10:07:06","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/project\/parser\/"},"modified":"2016-07-13T12:38:54","modified_gmt":"2016-07-13T19:38:54","slug":"parser","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/parser\/","title":{"rendered":"Parser"},"content":{"rendered":"

Definition<\/h2>\n
\n

The Arabic Parser determines the grammatical structure of Arabic sentences, such as which groups of words combine to form phrases and which words are the subject or the object of a verb. The Parser relies heavily on the Part-of-Speech (POS) Tagger to identify the correct part of speech for each word in an input sentence, and the Named-Entity Recognizer to identify named entities after it has been corrected using the Auto-Corrector.<\/p>\n

Features<\/h2>\n

Part-of-Speech disambiguation<\/h3>\n

The Parser relies on the\u00a0POS Tagger to resolve part-of-speech ambiguity.<\/p>\n

Named entity detection<\/h3>\n

The Parser also detects named entities in an Arabic sentence.<\/p>\n

Handling common Arabic mistakes<\/h3>\n

The Parser performs auto correction for common Arabic mistakes without further user effort.<\/p>\n

APIs<\/h2>\n