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

Hypernym and hyponym relation identification method based on gradient enhancement decision tree

A relationship recognition and decision tree technology, applied in character and pattern recognition, instrument, text database query, etc., can solve problems such as semantic drift, and achieve the effect of good hypernym relationship

Active Publication Date: 2020-04-07
ZHEJIANG UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But this method has the problem of "semantic drift" after many iterations

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
  • Hypernym and hyponym relation identification method based on gradient enhancement decision tree
  • Hypernym and hyponym relation identification method based on gradient enhancement decision tree
  • Hypernym and hyponym relation identification method based on gradient enhancement decision tree

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0043] Example: such as figure 1 As shown, a method for identifying the relationship between hyponyms and hyponyms based on gradient enhanced decision trees includes the following steps:

[0044] (1) The construction of random misplaced sample training set, such as figure 2 As shown, the details are as follows:

[0045]First, the corpus text is segmented through the lexical analysis system based on AliWS (abbreviation of Alibaba Word Segmenter). Then extract the hyponym pairs from the existing thesaurus for matching, and combine the text between the word pairs to construct positive samples. Displace the hypernyms of the successfully matched word pairs as negative sample word pairs. Then use the misplaced word to match the text to construct a random misplaced negative sample, such as:

[0046] (1)

[0047] (2)

[0048] (3)

[0049] After dislocation, , becomes . Then search for matching sentence paths in the corpus. After filtering, you get such as:

[0050] (1)...

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 relates to a hypernym and hyponym relationship identification method based on a gradient enhancement decision tree, which models the method into two types of tasks to determine whether an entity pair is a hypernym and hyponym entity relationship. In order to train a classification model, entity pairs and path information thereof are input, and the output is 1 (for the up-down relationship) or 0 (for the absence of the up-down relationship). A high-confidence recommendation set based on a positive classification result is obtained by jointly training two classifiers. A model quickly adapts to a rule mode of an unmarked corpus text by continuously iterating a high confidence set. The hypernym and hyponym relations of the e-commerce domain can be better mined.

Description

technical field [0001] The invention relates to the technical field of natural language processing, in particular to a method for identifying hyponym relations based on a gradient enhanced decision tree. Background technique [0002] The automatic mining and verification of the hypernymy relationship between entities is an important task in electronic commerce. The hyponym relationship expresses the relationship between a generic entity (hypernym) and a specific instance of it (hyponym). Examples include appliances and refrigerators. In e-commerce, mining this hyponym relationship can help to better understand user queries and product recommendations. [0003] However, in e-commerce, this task faces many challenges. First, the text corpus on the Internet often contains a lot of noise, and the text is updated frequently. Noise makes it difficult for general methods to obtain effective information from e-commerce texts. The high frequency of updates makes a lot of labor c...

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): G06F16/33G06F16/35G06F40/289G06K9/62
CPCG06F16/355G06F16/3344G06F18/2148
Inventor 潘翔阮义彰
Owner ZHEJIANG UNIV OF TECH