Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Semantic role labeling method and device

A semantic role labeling and semantic representation technology, applied in the field of semantic role labeling methods and devices, can solve the problem of high cost and achieve the effect of reducing dimensions

Active Publication Date: 2017-10-03
BEIJING BAIDU NETCOM SCI & TECH CO LTD
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] (3) More than 90% of the time is spent on constructing sparse features, looking up tables, and calling classifiers, which is very costly

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Semantic role labeling method and device
  • Semantic role labeling method and device
  • Semantic role labeling method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0029] figure 1 It is a schematic flowchart of a semantic role labeling method provided in Embodiment 1 of the present invention. This embodiment is applicable to the situation where semantic role labeling is performed on sentences in application scenarios where semantic role labeling of sentences is required, such as automatic summary generation, knowledge mining, sentiment analysis, statistical machine translation, or search correlation calculation. The method can be executed by a semantic role tagging device, which is implemented by software and can be built into terminal devices such as smart phones, tablet computers, notebook computers, desktop computers or personal digital assistants. see figure 1 , the semantic role labeling method provided in this embodiment specifically includes the following operations:

[0030] Operation 110. Obtain at least one classification feature of the word segmentation in the target sentence to be labeled.

[0031] Operation 120, determine...

Embodiment 2

[0052] Figure 2A It is a schematic flowchart of a semantic role labeling method provided in Embodiment 2 of the present invention. On the basis of the first embodiment above, this embodiment further optimizes the task of "identifying the predicate in the target sentence" among the three tasks of semantic role labeling. see Figure 2A , the semantic role labeling method provided in this embodiment specifically includes the following operations:

[0053] Operation 210. Obtain at least one classification feature of the word segmentation in the target sentence to be labeled.

[0054] Operation 220, determine the acquired semantic representation information of each classification feature.

[0055] In operation 230, the semantic representation information of each classification feature is used as an input of a pre-generated first neural network classifier, and the first neural network classifier is used to identify whether the word segmentation is a predicate.

[0056] In this ...

Embodiment 3

[0093] Figure 3A It is a schematic flowchart of a semantic role labeling method provided in Embodiment 3 of the present invention. On the basis of the first embodiment above, this embodiment further optimizes the task of "identifying the semantic case of the predicate" among the three tasks of semantic role labeling. see Figure 3A, the semantic role labeling method provided in this embodiment specifically includes the following operations:

[0094] Operation 310. Obtain at least one classification feature of the predicate in the target sentence to be labeled.

[0095] Operation 320, determine the acquired semantic representation information of each classification feature.

[0096] In operation 330, the semantic representation information of each classification feature is used as an input of the pre-generated second neural network classifier, and the second neural network classifier is used to identify the semantic lattice of the predicate.

[0097] In this embodiment, th...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The embodiment of the invention discloses a semantic role labeling method and device. Wherein, the method includes: obtaining at least one classification feature of the word segmentation in the target sentence to be marked; determining the semantic representation information of each classification feature obtained; using the semantic representation of each classification feature as the input of the pre-generated neural network classifier , using the neural network classifier to perform semantic role labeling on the word segmentation. The technical solution provided by the embodiments of the present invention can simply map complex and sparse features based on multiple words, multiple parts of speech, multiple dependent arc labels, and multiple dependent paths into dense features, thereby reducing the dimensionality and complexity of the feature space. The complexity of feature construction, and the combination of multiple features can be automatically realized.

Description

technical field [0001] The embodiments of the present invention relate to the field of computer technology, and in particular to a semantic role labeling method and device. Background technique [0002] Semantic role labeling, as one of the mainstream methods for analyzing the semantic backbone of sentences, focuses on describing the structural information of sentences from a semantic perspective. important application value. [0003] At present, the input of the system for semantic role labeling is usually the sentence to be labeled, and the output is the semantic structure tree of the sentence. Among them, the semantic structure tree describes all the semantic roles of the predicates in the sentence and the category of each semantic role. In the prior art, after receiving a certain sentence, the system usually implements the semantic role labeling of the sentence through the following scheme: first extract the words, parts of speech, dependency arcs, dependency paths, an...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F17/27G06F17/30
Inventor 吴先超
Owner BEIJING BAIDU NETCOM SCI & TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com