Unlock instant, AI-driven research and patent intelligence for your innovation.

Relation extraction method based on enhanced learning

A technology of relation extraction and reinforcement learning, which is applied in the field of relation extraction based on reinforcement learning, and can solve problems such as the inability of the relation extraction model to make correct predictions

Active Publication Date: 2019-07-26
SOUTHEAST UNIV
View PDF3 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In this case, the relation extraction model cannot make correct predictions without sufficient background information such as the type of entity

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
  • Relation extraction method based on enhanced learning
  • Relation extraction method based on enhanced learning
  • Relation extraction method based on enhanced learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0045] The technical solutions of the present invention will be described in detail below with reference to the accompanying drawings.

[0046] A method for relation extraction based on reinforcement learning according to an embodiment of the present invention includes:

[0047] Step 10) construct the relation extraction framework of reinforcement learning; Described relation extraction framework comprises relation extraction model based on DNN model, soft rules for representing relational human knowledge form and relational evidence containing query questions, external knowledge sources and intelligence. body;

[0048] Step 20) obtain the extraction result of the relation extraction model based on the DNN model;

[0049] Step 30) In the reinforcement learning environment, the agent dynamically adjusts the extraction result by using the soft rule and the relationship evidence.

[0050] For a relationship, soft rules are used to indicate whether a sentence directly expresses ...

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 invention discloses a relation extraction method based on enhanced learning. The method comprises the following steps: step 10) constructing a relation extraction framework of reinforcement learning; wherein the relationship extraction framework comprises a relationship extraction model based on a DNN model, a soft rule used for representing a relationship in a human knowledge form, relationship evidence containing a query problem, an external knowledge source and an agent; step 20) obtaining an extraction result of the relation extraction model based on the DNN model; and step 30) in theenhanced learning environment, enabling an intelligent agent to dynamically adjust the extraction result by using a soft rule and relations evidence, and the relations extraction method based on enhanced learning can enhance an existing relations extraction model based on DNN.

Description

technical field [0001] The invention belongs to the field of computer natural language processing, and in particular relates to a relation extraction method based on reinforcement learning. Background technique [0002] The purpose of relation extraction (RE) is to extract the semantic relations of entity pairs in text. For example, in figure 1 Given sentence #1 in , the goal of relation extraction is to determine the existence of the relation director_of_organization <e1, e2> between Phil Schiller and the sale department. Relation extraction is widely used in subsequent applications such as ontology construction, knowledge base (KB) construction, and question answering systems. To solve the problem of relation extraction, predecessors have done a lot of work, among which the use of deep neural network (DNN)-based models has gradually become the mainstream. These DNN-based models provide great power for learning features from large amounts of data, and significantly...

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
IPC IPC(8): G06F17/27G06F16/36
CPCG06F16/36G06F40/295Y02D10/00
Inventor 刘兵漆桂林柏超宇
Owner SOUTHEAST UNIV