Functions

NiuParser system involves seven text analysis technologies, that is, Word Segmentation , POS Tagging , Named Entity Recognition, Chunking, Constituent Parsing, Dependency Parsing and  Semantic Role Labeling

Word Segmentation
 Word Segmentation: Dividing a Chinese sentence into its component words. Input a Chinese sentence into the system and output its word segmentation result, all the words delimited by space. For example:

Input:欢迎使用我们的中文语言平台。
Output:欢迎 使用 我们 的 中文 语言 平台 。

POS Tagging
 POS Tagging: Input the word-segmented sentence into the system and output the sentence with all the words tagged with their categories. For example:

Input:欢迎    使用    我们    的    中文    语言   平台    。
Output:欢迎/VV 使用/VV 我们/PN 的/DEG 中文/NN 语言/NN 平台/NN 。/PN

Named Entity Recognition
 Named Entity Recognition: Input the word-segmented and -tagged sentence into the system, and output the named entities recognized in the sentence. For Example:

Input:欢迎/VV   使用/VV   我们/PN   的/DEG   中文/NN   语言/NN   平台/NN   。/PN
Output:

Chunking
 Chunking: Input the word-segmented and -tagged sentence, and output the chunking (basic phrases) recognized in the sentence. For Example:

Input:欢迎/VV   使用/VV   我们/PN   的/DEG   中文/NN   语言/NN   平台/NN   。/PN
Output:

Constituent Parsing
 Constituent Parsing: Input the word-segmented and -tagged sentence into the system, and output the syntactic structure of the constituents in the sentence (also called phrase structure). For example:

Input:欢迎/VV   使用/VV   我们/PN   的/DEG   中文/NN   语言/NN   平台/NN   。/PN
Output:

Dependency Parsing
 Dependency Parsing: Given a sentence that has been word-segmented and tagged, recognize the dependency relation between the words and expressions. For example:

Input:欢迎/VV  使用/VV  我们/PN  的/DEG  中文/NN  语言/NN  平台/NN  。/PN
Output:

Semantic Role Labeling
 Semantic Role Labeling: Input a constituent syntax tree, and recognize the semantic arguments of a specified predicate. Semantic Role Labeling is used for semantic structure that is abstract and independent of syntactic structure. For example:

Input:

Output: