During and after my PDF at the University of British Columbia (UBC), Canada I supervised, assisted, taught and helped CB, a then master student of EE. However, she did not give any credits to me in her thesis. My reports to her supervisor, and follow ups with UBC officials ended up nowhere!
You can download her thesis “Reliability analysis of discrete fracture network projections from borehole to shaft scale discontinuity data” from UBC website, e.g., the following link.
In the following, some memos, messages and other stuff related to the case are presented. These exemplify my contributions in her research and thesis.
Collaboration / Supervision / Assistance
In the beginning of my PDF and after some early chats with CB (one of many whom I helped) I felt a need to teach her how the fitting points works and to help her to understand the potential pitfalls; so I developed an interactive application in Java that demonstrates the ineffectiveness of R2 with respect to spatial scattering pattern of points. I gave the code to her and taught her how to play with it. In the app, she could easily move every point around to create different spatial patterns and see the resulting R2 value live.
The above image of the app is dated 2014-11-26 in my archive.
In the coming months, we met many times in the lab (several times a week) to discuss her research challenges, to teach her the multivariate statistics and DFNE, and to help her to analyze the resulting graphs. Aside from daily tips, I introduced to her the box-plot concept, how to create it, and also explained her how that would help her to summarize her many outputs. The simple concept of box-plot alone later turned out to be an important part of her thesis.
She had no prior knowledge of DFNE; once during our discussions I noticed a serious misconception in generating DFN (inferring parameters) while she was explaining the use of a closed-source commercial software in her work. She seemed following “click a button, next” approach without knowing the idea behind. A closed-source software and or lack of supervision may have played roles too. Anyways, I made a full discussion on that with examples.
She was doing simulations; and to be able to use the results she wrote to me:
Good Morning Younes,
For the text files, I have a few exports. I need to pull out the same information from each except I had multiple joint data exported in one file for the boreholes (Depth there are five and for the Angled runs there are nine. I have made 2 slides in PowerPoint to try and show what I am thinking for the format of the data that is pulled out. Please let me know if this makes sense.
Thank you so much in advance for your help, it is greatly appreciated.
In response, I developed a VBA code (ExtractInfoFromFilesInFolder.xlsm) to do the above task for her perfectly. I also created a visual tutorial in PowerPoint (HowtoSetAndUseVBACode_ExtractInfoFromFilesInFolder.pptx) showing how to use the code. I also issued a license for her so she could legitimately and freely use my code. I gave her a full copy of the package: code, tutorial and license.
The above figure (dated 2015-05-22 in my archive) shows a snapshot of part of the VBA code with added markers and notes so she can easily run it to extract the required information from all data files (and folders). The code puts all extracted information inside the excel file in a well-organized manner. She reportedly used the code multiple times with no hassle.
In addition to her requests, any time in our discussions that I realized she needs help to properly understand a concept, besides my explanations I developed codes to teach her; for example how 3D fractures could be observed in boreholes as traces, and their flat projection. The following figure (dated 2015-06-21) could help her to realize challenges in getting required information from boreholes to setup a realistic DFN, of the main subjects of her thesis.
I also showed her how various setup of boreholes may play role in decreasing the uncertainty, to help her to get a better idea on how the fracture information could be transferred from multiple boreholes onto a shaft. I also generated and showed her other visualizations on various errors in getting the right length for fractures from boreholes.
The above figure (2015-06-20) demonstrate options and challenges in inferring fracture network information from a set of boreholes in the same domain.
In addition, I ran several simulations to evaluate the number of boreholes versus their length for effectiveness of getting DFN information. I made several full discussions on this with her in person in the shared lab.
The following figure shows a snapshot of part of list of my other developed codes for her research; the dates are clearly shown; the name of files reflects the purpose of the codes properly. It can also be seen that I have continued (in 2016) to help her in her research even after I finished my PDF at UBC in 2015.
Let’s check out what inside the “Discussion.zip” file is.
Well, a lot of figures of box-plots plus their bounding curves. Not only I taught her what box-plot is and how it can be helpful for her research, but also I generated a lot of box-plots for her data as per her request. Once she learned, it could help her to significantly reduce the time and labor spent on analyzing and comparing the data / results.
The name of the files reflects CB’s data for easier match. Moreover, the file “inout.zip” includes even more simulations.
I gave her all the results (some examples shown above) in a variety of suitable data format, presentations, codes and so on.
I was the most qualified expert in the team on DFNE which is of the main topics of CB’s master research and thesis. My other expertise in multivariate and spatial statistics was the only reliable source for CB in dealing with data and interpreting the results correctly. Moreover, due to working in the same shared lab with her, I spent more times in assisting / supervising her than her supervisors combined.
CB failed to give credits to me. There is no acknowledgment in her thesis for my contributions to her research.
This post is aimed to help publicly the academia to preserve and defend its founding stone, the integrity.