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

Information retrieval method based on integrated support vector machine ranking

An information retrieval and support vector technology, applied in the field of information retrieval, can solve the problems of difficult training, increased training time, and low efficiency of algorithm training, and achieve the effect of improving the average accuracy

Inactive Publication Date: 2011-01-26
XIDIAN UNIV
View PDF2 Cites 17 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method still has the following problems: 1. The training efficiency of the algorithm is low, and it is difficult to carry out training when the training data set is too large; 2. Multiple trainings are required for parameter selection, which greatly increases the training time; relatively 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
  • Information retrieval method based on integrated support vector machine ranking
  • Information retrieval method based on integrated support vector machine ranking
  • Information retrieval method based on integrated support vector machine ranking

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0018] refer to figure 1 , the implementation steps of the present invention include as follows:

[0019] Step 1, set the training sample set

[0020] Set the training sample set Where m represents the total number of query objects, Denotes the feature vectors of all documents associated with the i-th query object, (j=1, 2, ..., n (i) ) represents the feature vector of the jth associated file, n (i) Indicates the total number of files associated with the i-th query object, means with x (i) The corresponding tag sequence, Γ is a matrix of M*l, where M=n (1) +n (2) +,...,+n (m) is the number of rows of the matrix, l represents the number of columns, the first column is the label column, the second column is the query object number, and the rest are the characteristics of the associated file, from the first row to the nth row (1) The row is the feature vector and label of the first query object associated file, that is (x (1) ,y (1) ), the nth (1) +1 row to nth ...

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 information retrieval method based on integrated support vector machine (SVM) ranking, mainly solving the problems of low training efficiency and ranking accuracy of the existing methods. The method comprises the following steps: (1) carrying out model training on training samples respectively according to different query objects to obtain an initial model group; (2) allocating ranking grades for a verification set by the ranking algorithm in the Ranking SVM algorithm according to each model in the model group and selecting the models making contributions to the average accuracy of the ranking grades after integration to form a system model set; and (3) each model in the system model set allocating the ranking grades for each eigenvector in a test set and using the sum of the ranking grades corresponding to the same eigenvector as output. The method improves the model training efficiency and ranking accuracy of the rank learning method for information retrieval, ensures the rank learning method to have universality and can be applied to network search engines, military intelligence retrieval and machine translation.

Description

technical field [0001] The invention belongs to the field of information retrieval, and relates to a machine learning method, in particular to a sorting learning method. Applicable to information retrieval. Background technique [0002] Information retrieval originated from the library's reference consultation and abstract indexing work. With the advent of the world's first electronic computer in 1946, computer technology gradually entered the field of information retrieval and was closely combined with information retrieval theory. Offline batch intelligence retrieval System and online real-time information retrieval system have been successfully developed and commercialized. With the continuous advancement of science and technology, driven by information processing technology, communication technology, computer and database technology, information retrieval has developed rapidly in various fields such as education, military affairs and commerce, and has been widely used. ...

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): G06F17/30
Inventor 郑喆坤沈彦波焦李成郑小皇张莉王爽马文萍尚荣华公茂果
Owner XIDIAN 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