Taozheng Zhang

Communication University of China

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Jianping Chai

Communication University of China

Min Zhao

Shenzhen Academy of Robotics

Zhenlong Zhang

Yamaguchi University

Binbin Wan

Communication University of China

Jiajian Meng

Communication University of China

Xiaojie Wang

Beijing University of Posts and Telecommunications

Jiaqi Guo

University of Electronic Science and Technology of China

Min Zhao

Wuhan University of Science and Technology

Saike He

Institute of Automation

Xue Bai

University of Chinese Academy of Sciences

Cooperative Institutions

Communication University of China

Chinese Academy of Sciences

University of Chinese Academy of Sciences

Beijing University of Posts and Telecommunications

Wuhan University of Science and Technology

Institute of Automation

Jiangsu University

Tianjin University

Shandong University

Xi'an University of Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Key Technologies of TV Programs Subjective Evaluations Based on Opinion Mining

Taozheng Zhang

This paper mainly introduces three key steps and technologies used in the processes of subjective evaluations of TV programs. Firstly, Chinese 2-POS subjective models are used to judge subjective and objective sentences. Secondly, feature words and sentiment words are extracted based on the templates of POS-Path. Thirdly, SO-PMI algorithm is used to determine and calculate the orientations of sentiment words. It effectively overcomes the three bottlenecks in the traditional subjective evaluations of TV programs.

Sentiment Analysis

Feature (linguistics)

10.1109/iscid.2017.99

Cite

Citations (0)

Based on SO-PMI Algorithm to Discriminate Sentimental Words' Polarity in TV Programs' Subjective Evaluation

Min Zhao Taozheng Zhang Jianping Chai

With the increasing of comment data about television programs on Internet, using text mining technology to analyze the massive data so that it can make subjective evaluation about television programs. It can give decision support and improve competitiveness for each department in radio and TV industry. This paper mainly introduces how to using sentiment-oriented pointwise mutual information (SO-PMI) to judge sentimental words' polarity and calculate sentimental words' intensity. It lays a solid foundation for the further study of subjective evaluation about television programs.

Polarity (international relations)

Pointwise

Sentiment Analysis

Pointwise mutual information

Foundation (evidence)

10.1109/iscid.2015.86

Cite

Citations (5)

Judging Subjective and Objective Sentence Based on 2-POS Subjective Models

Min Zhao Taozheng Zhang Jianping Chai

With the constant increasing of the TV programs text evaluate data on Internet, using emotional tendentiousness analysis technology to analyze the mass text data will help the program producers, program promulgator, network operator and government department improve their competitiveness and make a better decision, accordingly improve the competitiveness of the broadcasting and TV. The paper is on the basis of part of speech combination model. We firstly use the Chi-square statistic method to extract 2-POS subjective models, then calculate sentence's subjective degree, finally judge the sentence as subjective or objective sentence by comparing the degree with the threshold.

Statistic

Degree (music)

Operator (biology)

Constant (computer programming)

10.1109/uic-atc-scalcom-cbdcom-iop.2015.122

Cite

Citations (1)

Multimodal Fake News Detection Based on Contrastive Learning and Data Augmentation

Chun‐Shan Wang Taozheng Zhang

Fake News

10.1109/icecai62591.2024.10675158

Cite

Citations (0)

Research on Chinese Sentiment Analysis Based on Bi-LSTM Networks

Taozheng Zhang Jiaqi Guo

Chinese sentiment analysis is a very important branch of natural language processing. It has been receiving much attention in recent years. The bidirectional long and short-term memory network (Bi-LSTM) model has been well applied in the field of sentiment analysis because of its own characteristics. This experiment hopes to further explore the performance and application of the Bi-LSTM model in sentiment analysis. There are three main steps in the experiment. First, the collected Chinese reviews are segmented and vectorized. Then, the Bi-LSTM is trained and tested. Finally, the sentiment analysis result is obtained. With the help of the hyper-parameter adjustment and the dropout mechanism, the evaluation indicators of the experimental model have reached about 89%. What's more, based on the same experimental environment and experimental data, this experiment tested the accuracy of CNN, LSTM, CNN_LSTM, and Bi-LSTM. In addition, the trained Bi-LSTM was used to analyze reviews from Taobao and JD.COM. The specific operation is to collect reviews on a certain product from Taobao and JD.COM to perform a specific analysis with the model. Then find the advantages and disadvantages of the model in practical applications, so that the model can continue to be improved.

Sentiment Analysis

Dropout (neural networks)

10.1109/icisfall51598.2021.9627460

Cite

Citations (2)

Stereo Vision System & 3D TV Program Production

Advances in intelligent systems research/Advances in Intelligent Systems Research (2014)

Zhenlong Zhang Taozheng Zhang Jianping Chai

Stereo vision is a strong-crossed subject, and it has been one of the most hot topics in the study of computer.In more than 40 years of development, it has formed its own method and theory.This paper systematically reviews the present situation of stereo vision research, briefly introduces stereo vision system and 3D TV program production, and analyzes the error controlling through the production.

Stereo cameras

Computer stereo vision

Machine Vision

Stereo image

10.2991/scict-14.2014.29

Cite

Citations (0)

STFGSM: Intelligent Image Classification Model Based on Swin Transformer and Fast Gradient Sign Method

Taozheng Zhang Jiajian Meng

The convolutional neural network is relied upon by the mainstream image classification model to be achieved, but the convolutional neural network itself has defects such as easy loss of data. At the same time, deep learning models are vulnerable to adversarial perturbations, resulting in a decline in model performance. In order to effectively solve the above problems, this paper presents STFGSM, an intelligent image classification model based on Swin Transformer and fast gradient sign method. The attention mechanism is utilized by the Swin Transformer to extract picture features, with the traditional convolution operation being replaced. The field of the image is enhanced and information loss is avoided by this. Furthermore, the anti-interference capability of the model is strengthened through adversarial training that uses adversarial samples generated via the fast gradient sign method algorithm. The experimental results show that the classification performance of STFGSM outperformed other mainstream image classification models, whose speed is faster and adaptability to adversarial samples is stronger. In the future, more complex adversarial training strategies can be introduced on the basis of the model or the model can be extended to tasks in other fields such as target detection and image generation.

Contextual image classification

Adaptability

10.1145/3613330.3613339

Cite

Citations (0)

Domain Adaptation in Alternating Structure Optimization (ASO) Algorithm

Taozheng Zhang Xiaojie Wang Hui Tong

Domain Adaptation

10.2316/p.2011.717-018

Cite

Citations (1)

Multi-Task Learning in Conditional Random Fields for Chunking in Shallow Semantic Parsing

Pacific Asia Conference on Language, Information, and Computation (2009)

Saike He Xiaojie Wang Dong Yuan Taozheng Zhang Xue Bai

Alternating Structure Optimization (ASO) is a recently proposed linear Multitask Learning algorithm. Although its effective has been verified in both semi-supervised as well as supervised methods, yet they necessitate taking external resource as a prerequisite. Therefore, feasibility of employing ASO to further improve the performance merely rests on the labeled data on hand proves to be a task deserving close scrutiny. Catering to this challenging while untapped problem, this paper presents a novel application of ASO to the subtask of Shallow Semantic Parsing: Chunking. Our experiments on Chinese Treebank 5.0 present promising result in chunk analysis, and the error rate is reduced by 5.72%, proposing a profound way to further improve the performance.

Treebank

Chunking (psychology)

Semantic role labeling

Word error rate

Supervised Learning

Source

Cite

Citations (1)

Dilated Convolutional Neural Network with Joint Training for Chinese Word Segmentation

Taozheng Zhang Ziyao Liao

The task of Chinese word segmentation (CWS) is to segment a continuous Chinese text sequence into individual word sequences according to certain rules. As the most basic step of Chinese natural language processing tasks, the study of CWS is of great significance. In this paper, based on the method of deep learning, the radical features were added when pretraining the character vectors, which has led to the improvement of the outcome. Dilated Convolutional Neural Network (DCNN) was trained and compared with the baseline Bi-LSTM to verify the segmentation ability, and Conditional Random Field (CRF) was applied with two neural network models above-mentioned respectively to confirm the best model. The experiment results show that the DCNN model is better than the widely used Bi-LSTM model in the performance. The joint training method was also applied to further improve the training outcome, and it comes out the improvement on F1 value of joint training with several corpora.

Sequence labeling

Text segmentation

10.1145/3460569.3460582

Cite

Citations (0)