Introduction: In recent years, Weibo has become one of the most commonly used social networking tools for people. Therefore, it is very valuable to mine information in Weibo. Because of the fast and convenient nature of pictures, publishing microblogs with pictures is a new trend. At present, most of Weibo's sentiment analysis studies have focused on the text and can no longer be applied. The use of machine learning technology to perform sentiment analysis on pictures is an important part of achieving advanced human-computer interaction. It is of great significance for human-computer interaction, human-computer interfaces, and intelligent computers. This has become the current pattern recognition, machine learning, and cognitive science. One of the hot research topics in the field of research.
Title: Weibo Image Sentiment Analysis Based on Multi-core Learning Integrating Text Information
Abstract: On Weibo, pictures are one of the most important ways to express user emotions. More and more people only publish pictures on Weibo, because the pictures have fast and convenient natural properties. It is a new trend to publish microblogging with only pictures. At present, most of Weibo's sentiment analysis studies focus on texts or use images as supplementary information in texts, so they are not applicable in this situation. Although some sentiment analysis methods for pictures have been proposed, most of them ignore the semantic gap between low-level visual features and high-order picture emotions, or require a large amount of text information in the training and reasoning stages. This paper presents a new type of sentiment analysis based on SimpleMKL. Specifically, text information as a rich emotional source data, we can use it to improve the ability of SimpleMKL classification pictures. Once we get the picture classifier, no text is needed when predicting other untagged images. The experimental results show that the method has significant effect on the data obtained by the crawler and the Sina Weibo tag data. We find that this method is not only better than some common methods, such as support vector machine, Naive Bayes, KNN, random tree, algorithm, etc. It is also superior to some advanced methods based on color, gradient, texture and other image features.
Keywords: sentiment analysis; microblog; picture sentiment; multi-core learning
The first author introduction:
Junxin Tan
Nanjing University, Department of Computer Science, Key Laboratory of New Software Technology.
Via PRICAI 2016
Paper original file download
Copper Tube Terminals Without Checking Hole
Our company specializes in the production and sales of all kinds of terminals, copper terminals, nose wire ears, cold pressed terminals, copper joints, but also according to customer requirements for customization and production, our raw materials are produced and sold by ourselves, we have their own raw materials processing plant, high purity T2 copper, quality and quantity, come to me to order it!
Copper Tube Terminals Without Checking Hole,Cable Lugs Insulating Crimp Terminal,Cable Connector Tinned Copper Ring Terminal,Tubular Cable Lugs Crimp Terminal
Taixing Longyi Terminals Co.,Ltd. , https://www.txlyterminals.com