An overview and application of Chinese dependent syntactic analysis

Syntactic analysis is a key problem in the field of natural language processing, if it can be effectively solved, on the one hand, it can verify the correctness and perfection of the corresponding tree library construction system. On the other hand, it can also directly serve for various upper-level applications, such as search engine user log analysis and keyword recognition, such as information extraction, automatic question and answer, Other natural language processing related tasks such as machine translation.

Phrase structure and dependency structure are the two most widely studied grammatical systems in the present syntactic analysis. Here we introduce the dependent structure.

　　
The dependency syntax reveals its syntactic structure by analyzing the dependencies between components within a language unit. It is argued that the core verb in the sentence is the central component of the other component, and that it is not subject to any other component, and that all the dominant elements belong to the dominant person in some kind of dependency relationship.


1, only one ingredient in a sentence is independent;
2. Other ingredients are directly dependent on a component;
3, No ingredient can be dependent on two or more than two components;
4, if a component is directly dependent on the B component, and the C component is between A and B in the sentence, then C or directly depends on B, or directly dependent on a and B of a component,
5, the central component of the left and right sides of the other components do not occur with each other.

The phenomenon of the mutual domination and domination, dependence and dependence of the sentence components is common in the Chinese vocabulary (synthetic language), phrases, sentences, compound until the level of the sentence group can be used independently of the language units, this feature is the universality of the dependence of the syntactic analysis can reflect the semantic modification of the relationship between the elements of the sentence, It can obtain long-distance collocation information and is independent of the physical location of the sentence composition.

Attached: LTP dependency token meaning used by the dependency Analysis Module (http://www.ltp-cloud.com/demo/)

Now the dependency syntax analysis annotation relationship (total 14 kinds) and the meanings are as follows:

Relationship Type	Tag	Description	Example
Principal-predicate relationship	SBV	Subject-verb	I gave her a bouquet of flowers (I <--sent)
Dynamic-Guest relations	Vob	Direct object, Verb-object	I gave her a bouquet of flowers (send-to-flower)
Guest relations	IOB	Indirect object, Indirect-object	I gave her a bouquet of flowers (to her)
Front Object	Fob	Front object, Fronting-object	He reads all the books (books <--read)
Concurrent language	DBL	Double	He invited me to dinner (please--i)
Fixed-Chinese relations	ATT	Attribute	Red apple (red <--apple)
Structure in shape	ADV	Adverbial	Very beautiful (very <--beautiful)
Dynamic complement structure	Cmp	Complement	Finished the homework (done and finished)
Tie relationship	COO	Coordinate	Mountains and Sea (Mt. and the Sea)
Guest relations	Pob	Preposition-object	Within the trade zone (---within)
Left Attach relationship	LAD	Left adjunct	Mountains and sea (and <--Sea)
Right Attach relationship	Rad	Right adjunct	Kids (Kids)
Independent structure	Is	Independent structure	Two sentences are structurally independent of each other.
Core relationships	HED	Head	The core of the whole sentence

Previously more complex dependencies marked the meaning of the version:

Dependency token (24 total)
Fixed-Chinese relationship att (attribute)
Quantity relationship Qun (quantity)
Relationship COO (coordinate)
The same-bit relationship app (appositive)
Front additional relationship lad (left adjunct)
Post-Attach relationship rad (right adjunct)
Dynamic-Bingo Relationship vob (Verb-object)
POB relationship (Preposition-object)
Principal-predicate Relationship SBV (subject-verb)
Analogy relationship sim (similarity)
Core Hed (head)
Continuous structure VV (VERB-VERB)
Association structure CNJ (conjunctive)
Voice Structure MT (Mood-tense)
Independent structure is (independent structure)
Medium Structure ADV (adverbial)
Dynamic complement structure CMP (complement)
"The" Word Structure de
"Ground" word structure di
"Get" word Structure dei
"Put" the word structure BA
"By" Word Structure bei
Independent clause IC (independent clause)
Dependent clause DC (dependent clause)

Reference links

An overview and application of Chinese dependent syntactic analysis