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

Intention data hierarchical clustering method and system based on BERT

A technology of data hierarchy and clustering method, applied in the medical field, can solve problems such as the inability to quickly locate

Pending Publication Date: 2020-05-19
XIAMEN KUAISHANGTONG TECH CORP LTD
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In order to solve the problem that the current method for determining intentions is to manually look at historical dialogue data and ask professionals, and cannot achieve rapid positioning, the present invention proposes a hierarchical clustering method for intention data based on BERT, which includes the following steps:

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
  • Intention data hierarchical clustering method and system based on BERT
  • Intention data hierarchical clustering method and system based on BERT
  • Intention data hierarchical clustering method and system based on BERT

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0022] The main point of this solution is the language model, and the quality of the language model determines the effect of clustering. The role of the language model is to map the input sentence to a high-dimensional space. The three stages of language model development are one-hot, word2vec, and BERT. BERT outperforms word2vec in 11 NLP tasks, and even surpasses humans in individual tasks. Therefore, it is a good choice to use BERT as a language model.

[0023] The reason for choosing hierarchical clustering is that when the robot is initially made, only a small number of visitors’ intentions are often determined. Compared with the intentions added later, these intentions are actually subdivided intentions. For example, visitors often inquire about the price, and then separate the intention of inquiring about the price. The visitor may ask whether it is expensive. The reply words of these two intentions are different. Hierarchical clustering can just solve the problem of "...

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 an intention data hierarchical clustering method based on BERT. The intention data hierarchical clustering method comprises the following steps: S1, determining data of visitors; S2, obtaining a vector of each piece of data through BERT; S3, clustering the vectors in the step S2 by using a hierarchical clustering algorithm; and S4, adjusting the parameters of the hierarchical clustering algorithm to enable a clustering result to meet the standard. Through the scheme of the invention, the intention of the visitor in a certain field can be quickly determined.

Description

technical field [0001] The present invention relates to the medical field and the field of data computing technology, in particular to a BERT-based intent data hierarchical clustering method and system. Background technique [0002] In the early days of robots in the medical field, it was often necessary to manually review a large amount of dialogue data to determine the common intentions of visitors, or to ask experts familiar with the industry. In this way, a serious problem will arise. In the early stage of building the robot, a lot of time and manpower will be spent on determining the visitor's intention. [0003] A language model is an abstract mathematical modeling of language based on the objective facts of language, and it is a corresponding relationship. The relationship between the language model and the objective facts of language is like the relationship between the abstract straight line and the concrete straight line in mathematics. BERT is one of the most ad...

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): G06K9/62
CPCG06F18/231
Inventor 陈鑫肖龙源蔡振华李稀敏刘晓葳谭玉坤
Owner XIAMEN KUAISHANGTONG TECH CORP LTD