Sentiment analysis-based recommendation method and device for mixed user rating information

A technology of information recommendation and sentiment analysis, which is applied in the direction of using information identifiers to retrieve Web data, other databases, and special data processing applications. It can solve problems such as information asymmetry, affecting the accuracy of recommendation results, and inaccurate evaluation information. Achieve accurate results and improve the effect of collaborative filtering methods

Active Publication Date: 2018-08-21
哈尔滨米兜科技有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The defect of the collaborative filtering method is that its implementation needs to analyze and process the user's evaluation information, but in fact, due to the huge amount of review information and product information on the e-commerce website, information asymmetry is caused; at the same time, it is easy to appear many product review information Very few, but there are many reviews of popular products, resulting in inaccurate evaluation information, which in turn affects the accuracy of recommendation results

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
  • Sentiment analysis-based recommendation method and device for mixed user rating information
  • Sentiment analysis-based recommendation method and device for mixed user rating information
  • Sentiment analysis-based recommendation method and device for mixed user rating information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0047] The embodiment of the present invention provides a mixed user rating information recommendation method based on sentiment analysis, see figure 1 , the recommended method includes the following steps:

[0048] 101: Perform word segmentation processing on the data text through the second-order Markov chain, and obtain the training set after word segmentation;

[0049] Wherein, step 101 is specifically:

[0050] According to the characteristics of Chinese word segmentation, the existing Chinese text word segmentation methods can be divided into three categories: the first type is word segmentation method based on string matching; the second type is word segmentation method based on understanding; the third type is word segmentation method based on statistics method. The present invention uses the first word segmentation method based on character string matching. This method combines the second-order Markov chain to perform word segmentation processing on data documents. ...

Embodiment 2

[0062] The embodiment of the present invention combines the specific calculation formula to describe the scheme in embodiment 1 in detail, and the detailed description is as follows:

[0063] 201: Establish a word generation model;

[0064] In the process of text sentiment analysis, the word segmentation processing of the data text must be carried out first, and the word generation model is used in this process. The word generation model is shown in formula (1):

[0065]

[0066] Among them, W Seq* is the word sequence newly generated by the word generation model; W Seq≡ω 1 m =[ω 1 ,ω 2 ,.ω i ..ω m ] represents a set of sequences containing m words, where "≡" is equal to; ω i Indicates the i-th word; c 1 n Represents a sentence containing n words; for when c 1 n When it must appear, the probability value of W Seq appears; when c 1 n necessarily occurs when one of the W Seqs makes reaches the maximum value, the W Seq at this time is

[0067] For example: s...

Embodiment 3

[0099] The following is combined with specific tests, evaluation criteria, figure 2 and image 3 To verify the feasibility of the mixed user rating information recommendation method based on sentiment analysis provided by the present invention, see the following description for details:

[0100]In the experiment, when the number K of the nearest neighbor users is equal to 10, 20, and 30, the value of the weight factor α in the formula (6) is constantly changed, so that the product recommendation algorithm based on sentiment analysis can obtain the best value. Good predictive scoring accuracy.

[0101] Simultaneously, the present invention uses mean absolute error (MAE) as the evaluation standard of predictive score, compares with mean error, because mean absolute error is transformed into absolute value by deviation, the situation of positive and negative phase offset can not occur, therefore, mean absolute error It can better reflect the actual situation of the forecast er...

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 present invention discloses an emotional analysis based mixed user scoring information recommendation method and apparatus. The mixed user scoring information recommendation method comprises the following steps of calculating a similarity degree by means of a commodity mutually scored by users so as to obtain a user scoring similarity degree; calculating an emotional similarity degree by means of an emotional tendency and an emotional similarity degree of a text; obtaining a comprehensive similarity degree by combining the emotional similarity degree and the user scoring similarity degree; and acquiring a predictive score of the users for the commodity by means of the comprehensive similarity degree. The mixed user scoring information recommendation apparatus comprises a first acquisition module, a second acquisition module, a third acquisition module and a fourth acquisition module. According to the emotional analysis based mixed user scoring information recommendation method and apparatus, after emotional analysis on a user comment is added, a more accurate predictive score can be obtained, so that the aim of improving a conventional collaborative filtering method is fulfilled. As a final experiment result shows, a more accurate result is obtained in the method than a collaborative filtering recommendation algorithm based on users.

Description

technical field [0001] The invention relates to the fields of data mining, natural language processing and information retrieval, in particular to a method for recommending mixed user rating information based on sentiment analysis and a recommending device thereof. Background technique [0002] Currently, in related technologies of recommendation methods, the recommendation methods mainly include: a recommendation method based on collaborative filtering. The effect of this method is better, and it is one of the most widely used methods at present; it clusters users by calculating the similarity of users, and users with the same cluster label have similar interests. products laid the foundation. For example: a certain user in the tennis group purchases a tennis racket of a certain famous brand, and recommends the tennis racket to other users in the tennis group. Due to the good performance of the collaborative filtering method, the recommendation system based on this method...

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 Patents(China)
IPC IPC(8): G06F17/30
CPCG06F16/35G06F16/955
Inventor 喻梅成基元于健高洁徐天一盛杰
Owner 哈尔滨米兜科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap