Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Subjective and objective classifier building method and system

A subjective and objective classifier technology, applied in the field of subjective and objective classifier construction, can solve the problem of low classification accuracy of subjective and objective classifiers, and achieve the effect of improving performance and accuracy

Active Publication Date: 2015-01-07
SUZHOU UNIV
View PDF5 Cites 17 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] It can be seen that the construction of existing subjective and objective classifiers only focuses on the training of questions. However, the categories of some questions may have ambiguities, that is, they may belong to different subjective and objective categories due to different answers, resulting in the final training results. The classification accuracy of the subjective and objective classifier is low

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
  • Subjective and objective classifier building method and system
  • Subjective and objective classifier building method and system
  • Subjective and objective classifier building method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0064] Embodiment 1 discloses a method for constructing a subjective and objective classifier, refer to figure 1 , the method may include the following steps:

[0065] S101: Perform subjective and objective classification training on the first preset classification algorithm by using a predetermined number of question training samples to obtain a question-based classifier.

[0066] refer to figure 2 , before carrying out each step of the inventive method, need at first carry out following pretreatment process:

[0067] S201: Grab a first preset number of question-answer pairs;

[0068] S202: Filter out question-answer pairs with low reference value, wherein the question-answer pairs with low reference value include question-answer pairs of subjective and objective categories that cannot be manually distinguished from questions or answers, and question-answer pairs with irrelevant answers and invalid answers;

[0069] S203: Manually mark the subjective and objective categor...

Embodiment 2

[0106] In the second embodiment, refer to image 3 , the subjective and objective classifier construction method may also include the following steps:

[0107] S104: Use the test sample to verify the classification accuracy of the target subjective and objective classifier, and evaluate the classification performance of the target subjective and objective classifier based on the classification accuracy.

[0108] Among them, the verification and evaluation process in this step includes:

[0109] Subjectively and objectively classifying the test samples by using the target subjective and objective classifier;

[0110] Comparing the classification categories of the test samples classified by the target subjective and objective classifier with the label categories of the test samples, and obtaining the number n of test samples whose classification categories are the same as the label categories 1 ;

[0111] Based on the formula k=n 1 / n 0 , to obtain the classification accur...

Embodiment 3

[0118] Embodiment 3 discloses a system for constructing subjective and objective classifiers, and the system corresponds to the methods for constructing subjective and objective classifiers disclosed in Embodiments 1 and 2.

[0119] First, refer to Figure 5 , corresponding to Embodiment 1, the system includes a first training module 100 , a second training module 200 and a fusion module 300 .

[0120] The first training module 100 is configured to use a predetermined number of question training samples to perform subjective and objective classification training on the first preset classification algorithm to obtain a question-based classifier.

[0121] The second training module 200 is used to use the predetermined number of answer training samples to perform subjective and objective classification training on the second preset classification algorithm to obtain an answer-based classifier, wherein the answer training samples and the question training The samples are in one-t...

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 subjective and objective classifier building method and system. The subjective and objective classifier building method and system are characterized in that emphasis is put on training questions and answers, base classifiers are built in terms of the question text and the answer text and then are infused, and a final subjective and objective classifier is obtained. Therefore, the answer classification is added in the subjective and objective classification, the question classification is corrected and calibrated by combining the answer features, and therefore subjective and objective classification based on question and answer complementation is achieved, the shortcoming of low accuracy of the classifier caused by ambiguity of a question training sample is overcome, the accuracy of classifying questions by aid of the subjective and objective classifier is improved, and further the performance of a question and answer system is improved.

Description

technical field [0001] The invention belongs to the technical field of natural language processing and pattern recognition, and in particular relates to a method and system for constructing a subjective and objective classifier. Background technique [0002] Automatic question answering technology is a research hotspot in the field of natural language processing, which generally includes three main components: question classification, information retrieval and answer extraction. [0003] Question classification can effectively reduce the space of candidate answers and improve the accuracy of answers returned by automatic question answering systems. Among them, subjective and objective classification is the basis for realizing question classification. For the questions raised by users, the automatic question answering system first needs to classify them subjectively and objectively, and then further subdivide the categories of questions on the basis of subjective and objectiv...

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 Applications(China)
IPC IPC(8): G06F17/28G06F17/30
Inventor 李寿山张栋周国栋
Owner SUZHOU UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap 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