This article discusses how a prolific author in an established institution managed to produce over 200 articles within 3 years as an academic.
First we look at the following three publications:
These publications all have one thing in common: wind turbine
- Support vector regression methodology for wind turbine reaction torque prediction with power-split hydrostatic continuous variable transmission, Journal Name: Energy, Volume 67, 1 April 2014, Pages 623–630
- Wind turbine power coefficient estimation by soft computing methodologies: Comparative study, Energy Conversion and Management, Volume 81, May 2014, Pages 520–526
- Survey of the most influential parameters on the wind farm net present value (NPV) by adaptive neuro-fuzzy approach, Renewable and Sustainable Energy Reviews, Volume 57, May 2016, Pages 1270–1278
SUBJECT
The objective of Paper 1 was to capture maximum energy from the wind by predicting the optimal values of the wind turbine reaction torque. To build an effective prediction model, the authors have decided to use polynomial and radial basis function (RBF) and have applied them as the kernel function of Support Vector Regression (SVR) for prediction of wind turbine reaction torque. Instead of minimizing the observed training error, SVR_poly and SVR_rbf attempt to minimize the generalization error bound so as to achieve generalized performance. The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by their proposed approach. Results show that SVRs can serve as a promising alternative for existing prediction models.
Paper 2 is still on wind power and still using RBF dan SVR to estimate the optimal power coefficient value of the wind turbines. Instead of minimizing the observed training error, SVR_poly and SVR_rbf attempt to minimize the generalization error bound so as to achieve generalized performance. The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by the SVR approach in compare to other soft computing methodologies.
Nothing much was discussed about the data and how they were obtained! We are very certain the same data were used for both papers!
So what is the contribution of the second paper?As far as computer science is concerned, the findings of the second paper is the same as that of the first! What a waste of taxpayers money funding this research! What a waste of subscribers money buying journals that recycles facts and findings!
The main objective of Paper 3 is to maximize wind farm efficiency. The optimal by taking economic aspects into account. The net present value (NPV) is the most important criteria for project investment estimating. The general approach in deciding the distinctive choice for a task through NPV is to treat the money streams as known with conviction. Even little deviations from the decided beforehand values might effectively negate the choice. To assess the investment risk of wind power project, this paper constructed a process that selected the most influential wind farm parameters on the NPV with adaptive neuro-fuzzy (ANFIS) method. This procedure is typically called variable selection, which corresponds to finding a subset of the full set of recorded variables that exhibits good predictive abilities. Variable seeking utilizing the ANFIS system was performed to figure out how the seven wind farm parameters affect the NPV of the wind farm.
So there is a slight difference between Paper 3 and Papers 1 and 2.
Abstract from Paper 1
Nowadays the use of renewable energy including wind energy has risen dramatically. Because of the increasing development of wind power production, improvement of the prediction of wind turbine output energy using classical or intelligent methods is necessary. To optimize the power produced in a wind turbine, speed of the turbine should vary with wind speed. Variable speed operation of wind turbines presents certain advantages over constant speed operation. This paper has investigated power-split hydrostatic continuously variable transmission (CVT). The objective of this article was to capture maximum energy from the wind by prediction the optimal values of the wind turbine reaction torque. To build an effective prediction model, the polynomial and radial basis function (RBF) are applied as the kernel function of Support Vector Regression (SVR) for prediction of wind turbine reaction torque in this research study. Instead of minimizing the observed training error, SVR_poly and SVR_rbf attempt to minimize the generalization error bound so as to achieve generalized performance. The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by our proposed approach. Results show that SVRs can serve as a promising alternative for existing prediction models.
Abstract from Paper 2
Wind energy has become a large contender of traditional fossil fuel energy, particularly with the successful operation of multi-megawatt sized wind turbines. However, reasonable wind speed is not adequately sustainable everywhere to build an economical wind farm. In wind energy conversion systems, one of the operational problems is the changeability and fluctuation of wind. In most cases, wind speed can vacillate rapidly. Hence, quality of produced energy becomes an important problem in wind energy conversion plants. Several control techniques have been applied to improve the quality of power generated from wind turbines. In this study, the polynomial and radial basis function (RBF) are applied as the kernel function of support vector regression (SVR) to estimate optimal power coefficient value of the wind turbines. Instead of minimizing the observed training error, SVR_poly and SVR_rbf attempt to minimize the generalization error bound so as to achieve generalized performance. The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by the SVR approach in compare to other soft computing methodologies.
We tried very hard to see the difference in the two abstracts. Let’s give the authors the benefit of the doubt. Let’s look at their conclusions.
CONCLUSION FROM THE PAPERS
Paper 1:
The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by our proposed approach. Results show that SVR can serve as a promising alternative for existing prediction models. It can be seen from the experiment that the prediction model overcomes the main shortage of artificial neural network without defining net work structure and trapping in the local optimum.
Paper 1:
The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by our proposed approach. Results show that SVR can serve as a promising alternative for existing prediction models. It can be seen from the experiment that the prediction model overcomes the main shortage of artificial neural network without defining network structure and trapping in the local optimum.
The above paragraphs were extracted from the conclusion of the respective papers. Can anyone tell me the difference in the two conclusions? They are the same word for word!
Let’s still give the authors a chance. Let’s look at their similarity index.
Table 1: Similarity Index
| Paper | Similarity Index |
| 1 | 74% |
| 2 | 81% |
| 3 | 40% |
Paper 1 has a Similarity Index of 74%, Paper 2 81% while Paper 3, 40% (Please refer to Table 1). How these papers passed the stringent reviewing processes of the different journals baffles me.
As shown in Table 2a and 2b the contents of Paper 1 and Paper 2 overlaps with several papers written by the author and his team in 2014. These papers are from the Infra Red Physics & Technology, Applied Intelligence, Renewable and Sustainable Energy Reviews, Physica E, Solar Energy, Bulletin of Earthquake Engineering and the Journal of the Optical Society of America ( Pls refer to Table 2a, 2b and the Reference).
There is a 39% overlap of Paper 1 with [1] (Table 2a). The first author of Paper 1 is the co- author of [1], while the first author is an academic from Serbia. This looks like an international collaboration funded by University of Malaya (UM) as acknowledged in the paper (Grant CG043-2013). It is strange that the University of Malaya is funding this research on optic lens when none of the researchers are optic lens expert. In fact, one of the UM academician involved in [1] is a historian as stated in his UMExpert website and the others are computer scientists. The computing techniques applied here are artificial intelligence/machine learning techniques and none of these computer scientists are from AI.
Let me give another summary; this time taken from [1]
In this study, the polynomial and radial basis function (RBF) are applied as the kernel function of Support Vector Regression (SVR) to estimate and predict estimate MTF value of the actual optical system according to experimental tests. Instead of minimizing the observed training error, SVR_poly and SVR_rbf attempt to minimize the generalization error bound so as to achieve generalized performance. The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by the SVR_rbf approach in compare to SVR_poly soft computing methodology. accuracy and capability of generalization can be achieved by the SVR_rbf approach in compare to SVR_poly soft computing methodology.
Wait! Where have I seen this before? Oh yes, the abstracts from Paper 1 and Paper 2!
Again, the question remains of what is the difference between the various researches done by this group of authors as published in Papers 1,2 and [1]?
Table 2a also shows a 36% overlap between Paper 1 and [2] which is also funded by the University of Malaya, Malaysia, under the RP002D-13ICT research grant. The first author of this paper is the same author from Serbia, as in [1].
Paper 1 also have 33% overlap with [3], where this time the first author of Paper1 is the first author of [3] but this timeusing the job assignation of Islamic Azad University, Iran. Paper 1 also have a 33% overlap with [4] where a physicist from UM is a co-author and the research is funded by the University of Malaya under UM.C / 625/1 / HIR / KPT / SCI / 29, RP008E-13AET dan RU001-2014. Etc. etc. Please refer to Table 2a.
Paper 2 has a Similarity Index of 81%! With such a huge similarity index, what is the originality of Paper 2? At the risk of sounding like a parrot, what is the contribution of Paper 2? What was the purpose of publishing Paper 2? Please refer to Table 2b and the reference for details.
Tables 2a and 2b show that the first author of Paper 1 does not really write new articles to be published but merely recycles his articles. The details of the papers that match Papers 1 and 2 can be found in the references.
This is how the author is able to publish over 200 articles within 3 years or so.
We will not discuss the overlap of Paper 3, but wstill th a Similarity Index of 40%, the degree of overlap is relatively small. However, most journals will not consider any papers with a similarity index above 25% and we are surprised that this paper has been published .
Table 2a: Articles that match with Paper 1
| Artikel Lain | Percent Matched |
Table 2b: Articles that match with Paper 1
| Artikel Lain | Percent Matched |
Notice how Paper 2 overlaps with the same articles as in Paper 1. This just goes to prove that the author recycles his papers.
The conclusion taken from [2] :
The experimental results show that an improvement in predictive accuracy and capability of generalization can be achieved by our proposed approach. Results show that SVR can serve as a promising alternative for existing prediction models. It can be seen from the experiment that the prediction model overcomes the main shortage of artificial neural network without defining network structure and trapping in the local optimum.
Wait haven’t they come to this conclusion in two different papers? So, what is the contribution to knowledge from all these publications? Why are all these journals regurgitating facts and findings?
DISCUSSIONS
The three main papers in this discussion is on wind turbine. They are funded by the research grants of the University of Malaya. It is uncertain why the University of Malaya is allowing their researchers with grants to be involved in publication misconduct. Some of these researchers are professors. One tends to wonder if the papers have such a high percentage of duplication, if there was any research done at all. Is this merely a publication misconduct or the more serious research misconduct.
What is the contribution to knowledge from subsequent papers after 1 paper has been published?
REFERENCES:
[1] Dalibor Petkovica, Shahaboddin Shamshirbandb, Hadi Saboohic, Tan Fong Angd, Nor Badrul Anuard, Zulkanain Abdul Rahmane, Nenad T. Pavlovica, Evaluation of modulation transfer function of optical lens system by support vector regression methodologies – A comparative study, Infrared Physics & Technology, Volume 65, July 2014, Pages 94–102.
[2]Dalibor Petkovic,Shahaboddin Shamshirband,Hadi Saboohi,Tan Fong Ang, Nor Badrul Anuar, Nenad D. Pavlovic, Support vector regression methodology for prediction of input displacement of adaptive compliant robotic gripper, Applied Intelligence, October 2014, Volume 41, Issue 3, pp 887–896.
[3] Zeynab Ramedania, Mahmoud Omidb, Alireza Keyhanib, Shahaboddin Shamshirbandc, Benyamin Khoshnevisanb, Potential of radial basis function based support vector regression for global solar radiation prediction, Renewable and Sustainable Energy Reviews, Volume 39, November 2014, Pages 1005–101
[4] Rozalina Zakariaa, Siti Munirah Che Noh, Dalibor Petkovic, Shahaboddin Shamshirband, Richard Penny, Investigation of plasmonic studies on morphology of deposited silver thin films having different thicknesses by soft computing methodologies—A comparative study, Physica E: Low-dimensional Systems and Nanostructures, Volume 63, September 2014, Pages 317–323.
[5] Zeynab Ramedania, Mahmoud Omidb, Alireza Keyhanib, Benyamin Khoshnevisanb, Hadi Saboohic, A comparative study between fuzzy linear regression and support vector regression for global solar radiation prediction in Iran, Solar Energy, Volume 109, November 2014, Pages 135–143.
[6]Shatirah Akib, Sadia Rahman, Shahaboddin Shamshirband, Dalibor Petkovic, Soft computing methodologies for estimation of bridge girder forces with perforations under tsunami wave loading, Bulletin of Earthquake Engineering March 2015, Volume 13, Issue 3, pp 935–952.
[7] Rozalina Zakaria, Ong Yong Sheng, Kam Wern, Shahaboddin Shamshirband, Ainuddin Wahid Abdul Wahab, Dalibor Petkovic, and Hadi Saboohi, Examination of tapered plastic multimode fiber-based sensor performance with silver coating for different concentrations of calcium hypochlorite by soft computing methodologies — A comparative study, Journal of the Optical Society of America AVol. 31, Issue 5,pp. 1023-1030, (2014).
