The notice said:
Recently, the college received a phone call to report that Li Ruifeng, a graduate student in 2008, was suspected of plagiarism. The college attached great importance to this matter and set up an investigation team to verify the situation. After investigation, it is true that Li Ruifeng plagiarized his master's thesis.
According to the Academic Norms for Postgraduates of Tianjin University, the 8th Sub-committee of academic degree evaluation committee of Tianjin University held its103rd meeting on June 20th, 2008. At the meeting, the plagiarism of Li Ruifeng's master's degree thesis and its handling opinions were discussed, and the decision to revoke Li Ruifeng's original degree and withdraw the master's degree certificate awarded by Li Ruifeng was unanimously adopted, which has been reported to the school academic degree evaluation committee for approval.
Recently, some netizens said that a master's degree thesis of Tianjin University was suspected of plagiarizing a master's degree thesis of Inner Mongolia Agricultural University on a large scale.
These two papers are respectively the master's thesis of Li Ruifeng, a graduate of architectural and civil engineering from the School of Architecture and Engineering of Tianjin University in 2008, and the master's thesis of Wu Xinhui, a graduate of Agricultural Soil and Water Engineering from Inner Mongolia Agricultural University in 2005, entitled "Research on Strength Prediction of Ordinary Concrete Based on Artificial Neural Network" (hereinafter referred to as "Wu Xinhui's thesis").
Wu Xinhui's thesis was completed in May 2005 and Li Ruifeng's thesis was completed in May 2008. In terms of time, Li Ruifeng's thesis was finished three years later than Xin Hui's.
Thesis reporter downloaded the above two papers from China HowNet. After careful comparison, it is found that the two papers are highly similar in content and text, and there is no difference in the title, publication time and page number of 74 references.
It is worth noting that in the "thank you" part of Li Ruifeng's paper, he also thanked Wu Xinhui, who had become an associate professor at Inner Mongolia Agricultural University at that time.
In view of Li Ruifeng's alleged plagiarism of Wu Xinhui's paper and his gratitude to Wu Xinhui, the newspaper recently contacted Wu Xinhui for verification. Wu Xinhui, who is currently teaching in the Forestry College of Inner Mongolia Agricultural University, told the newspaper that she didn't know Li Ruifeng at all. Her master's thesis was original, and I don't know if it was copied by others.
After comparing the two papers, it is found that although the topics of the two papers are not exactly the same, they both focus on the prediction of concrete strength by neural network.
The Chinese abstracts of the two papers are very similar. The Chinese abstract of Wu Xinhui's thesis is divided into two natural paragraphs, namely:
"Compressive strength is one of the most important properties of concrete, the core content of concrete quality control, and an important basis for structural design and construction. The code stipulates that the concrete strength of structural members can only be obtained after 28 days of standard curing, which can not meet the time requirements of modern construction and may leave hidden dangers. Therefore, it is of great significance to develop the early rapid determination technology of concrete strength and improve the prediction accuracy.
Based on the research results of concrete strength prediction at home and abroad, combined with the basic principle of artificial neural network, the selection of input vectors, network structure, transfer function and other parameters of the network is studied by using MATLAB neural network toolbox. On this basis, three on-site substations in central and western Inner Mongolia are selected, and a large number of standard samples with the same maintenance conditions are made as training samples and test samples of the network model. The basic BP algorithm, adaptive learning rate algorithm with additional momentum factor and L-M algorithm are used to train the network respectively. After a lot of trial calculation and comparison of simulation results, an ordinary concrete strength prediction model with reasonable network structure, fast convergence speed and high accuracy is finally established by using L-M algorithm. Compared with the prediction results of multiple linear regression model, BP network model has higher accuracy, and the prediction error is controlled within 3%, which can greatly avoid the problem of large deviation of strength prediction in concrete construction. "
At the same time, it can improve the quality of power supply voltage and power supply reliability. See Table 4 for the meteorological data and temperature statistics of 30 years (197 1 to 2000) provided by the Meteorological Bureau of Urad Qianqi. In this study, 87 groups were sampled from the foundation and cable trench of the main transformers 1#, 2# and 3# of this station. "