"Move to strike writing expert. Eric Schupps is not an expert witness in writing analysis. It is not listed on his resume. He does not even mention writing analysis. He instead lists expertise in SharePoint software, beer and whiskey.
The software Schupps used to analyze the writings is seriously flawed. JGAAP is a free beta program which only does closed class attribution. As per the author “you could give it a copy of Moby Dick and ask it which Marx Brother wrote it, and it would have to pick one of the Marx Brothers.” The creators of the program stated "there are many key problems such as the open class problem, the adversarial problem, and the co-authorship problem. It is emerging technology. It's not proven." It does not meet Texas Rules of Evidence.
Their expert Schupps took supposed samples of Defendant’s writings which were not authenticated. He does not know who wrote them. He assumed Defendant wrote them when Defendant did not write all parts of all articles. There were multiple co-authors. The software cannot be used with co-authors. This is clearly a case of garbage in, garbage out. He does not meet Texas Rules of Evidence as an expert.
More from the author of the software in communication to the Defendant stated, “I would like to tell you that your assumptions are correct in that JGAAP is not intended as a final product to be used in court cases. JGAAP is built as a test bed for different authorship attribution techniques and meant to be used to compare their effectiveness at solving different problems. There are many different methods in JGAAP which work to varying degrees and I assume if you try enough of them you could find a handful that tell you whatever result you have preconceived. Now the design of the experiment you put forward negates any results since it is not even testing your writing.”
The expert and software program used do not meet the standards for Texas rules of evidence for experts and analysis. This analysis should be thrown out completely. This testimony is beyond the witnesses expertise and is inadmissible. He does not have the proper education, experience and training to have acquired competence to assist the fact finder as required by the Texas Rules of Evidence 702."
Just read through Eric Shupps of BinaryWave trial transcript. Here is my cross examination of Eric Shupps of BinaryWave, Sharepoint. The minutes make Eric Shupps look really bad. He comes across as totally crazy. I'm amazed he'd allow Amanda Lollar to post them. It is extremely clear he does not know what he is talking about. Eric Shupps admits he's not an expert. Eric Shupps admits he's never used the software before for a client. Eric Shupps also comes across as a want-to-appear-to-be-a-know-it-all. Eric Shupps uses basic words incorrectly. I think he's trying to appear "learned" by using bigger words but he doesn't know the meaning, such as "Google bomb." I think Eric Shupps has a chip on his shoulder because he didn't go to college, just like Amanda Lollar.
I googled for five minutes today. I saw that he just got a business loan for BinaryWave from Lighter Capital. Don't they realize that Eric Shupps went bankrupt? Eric Shupps declared BinaryWave to be worth nothing. Eric Shupps defaulted on a previous business loan. Why would anyone loan him money?
When I first looked at Eric Shupps @eshupps #sharepoint of BinaryWave's resume and website I felt something fishy. It seemed really fake. When I met him in court my feelings were confirmed. Eric Shupps basically just bullshitted about "Google bombs" when he didn't even know the basic definition. Eric Shupps also bullshitted about the authorship software. Eric Shupps’ presentation was laughable yet he kept insisting. My gut feeling was correct. I ran one records check and found a recent bankruptcy filing. Here it is, a public document. His SS# is not listed. Eric Shupps ran up a bunch of credit cards, bought two trucks with loans, took money out of his house, has an ex-wife, owes her child support, has a current wife, has three kids and is only 40. Eric Shupps admits in his filing that his company BinaryWave is worthless. Eric Shupps admits he barely makes any money, definitely not enough to support his "lifestyle" of fine beer and whiskey and traveling abroad. That explains his behavior in court. Eric Shupps acted like he knew that I knew he was a fraud. So many frauds. Bat experts, IT experts...
Eric Shupps stated that he had to take care of his dad who was dying of cancer. That doesn't jive with the bankruptcy docs or his Facebook page. He filed for BK October 2010. His dad died January 2012. Plus, you don't vacation around the world when your dad is dying. I didn't leave town at all when I took care of my grandmother. I only see a small medical bill for a child in the docs. Story teller just like his clients. How do these people live with themselves with all the stories they tell?
Amanda Lollar of Bat World Sanctuary and now Eric Shupps of Binarywave are libeling and defaming me. They are stating that I created a "Google Bomb" about Amanda Lollar, Bat World Sanctuary. Here is the true definition of a Google Bomb. "The terms Google bomb and Googlewashing refer to practices, such as creating large numbers of links, that cause a web page to have a high ranking for searches on unrelated or off topic keyword phrases, often for comical or satirical purposes. In contrast, search engine optimization is the practice of improving the search engine listings of web pages for relevant search terms."
Big difference between Google Bombs which are now impossible due to algorithm changes and proper search engine optimization. Eric Shupps would know that if he were a true SEO expert. Obviously he is not. Amanda Lollar on the other hand is trying to spam the search engines. Fortunately she doesn't know what she's doing.
Eric Shupps of Binary Wave is not an expert on authorship. Eric Shupps never testified as an expert in a trial or deposition. This was the first time he used this free beta software JGAAP for a client. The author of the software stated it has limitations and should not be used in a court of law. The author stated it is a closed system. You must give it the names of possible authors then it chooses the most likely. It does not identify the true author. It cannot be used with multiple authors for a document. The samples which they stated were samples of my writing were not 100% my writing. Some were other people's articles, some included huge passages of quotes from others. A few people edited my articles. Garbage in, garbage out.
From the transcript:
Eric Shupps: I'm sorry, Your Honor. I don't know if I'm allowed to ask questions.
Judge William Brigham: You're not. You're to respond to the questions Ms. Cummins asks you.
(BY MS. CUMMINS) Have you ever been an expert witness in deposition for authorship?
Eric Shupps: No.
Q. Have you ever been an expert witness for authorship before?
Eric Shupps: No.
Q. So this is your first time?
Eric Shupps: Yes.
Q. Well, then let me ask you a question: If you were to give this software program the possible authors of Groucho, Chico, and Harpo Marx and then you gave it a copy of Moby Dick, and asked them which Marx Brothers wrote of the Marx Brothers wrote it, would it have to pick one of the Marx brothers?
Eric Shupps: Yes, it would. That’s what closed class attribution is all about.
Q. So you just stated that you're not an expert?
Eric Shupps: I believe I have maintained that all along. Yes.
Q. Do you realize that this software is a beta software?
Eric Shupps: Of course, yes.
Q. Again, do you have some physical proof or evidence that can prove beyond a shadow of a doubt that I absolutely wrote and posted those articles on Indybay?
Eric Shupps: There is no evidence that would exist to that.
Q. A few last questions. You stated earlier that this is the first time you've been an expert witness in of law, correct?
Eric Shupps: Correct.
Q. How many other times have you done authorship analysis of online content? How many other reports have you done for clients for pay?
Eric Shupps: For clients for pay. None.
Q. So this is the first time you've done one of these for a client, and it's the first time you've been an expert witness in court, correct?
Eric Shupps: That is correct.
Q. And just so we're clear, are you 100 percent certain that I wrote all of the blogs in Animal Advocates that you used that you attributed to me?
Eric Shupps: No.
Q. Okay. So you -- have you ever heard of the term garbage in, garbage out?
Eric Shupps: I'm very familiar with that term.
____________________________
Eric Shupps admitted in court he had no proof that I wrote the articles in question in Exhibit 18. None of the items in Exhibit 18 are included in the court order.
Proof that Eric Shupps has no idea what he's doing. From the author of the software. First my email to them, May 27, 2012:
Can this program legally be used by expert writing witnesses for trial purposes? A supposed expert writing witness took some articles from my blog which I did not write. He attributed them to me. He then compared them to other articles on anonymous websites that I also did not write. He claimed your program showed with 71 to 95% certainty that I wrote the anonymous articles. I didn't even write the comparison ones! Garbage in, garbage out.
I just read in the evllabs FAQ page that stated the program will pick one known author. I was listed as a known author along with three professional male writers. I'm not a professional writer and didn't even write the samples. Is this software program sophisticated enough for trial purposes? This seems to be a major flaw for legal purposes, i.e. "JGAAP only does so-called "closed class" attribution, which is to say, it picks the most likely author from among the set of authors you tell it about." Thanks.
JGAAP's response, June 6, 2012:
I would like to tell you that your assumptions are correct in that JGAAP is not intended as a final product to be used in court cases. JGAAP is built as a test bed for different authorship attribution techniques and meant to be used to compare their effectiveness at solving different problems. There are many different methods in JGAAP which work to varying degrees and I assume if you try enough of them you could find a handful that tell you whatever result you have preconceived.
Now the design of the experiment you put forward negates any results since it is not even testing your writing.
From: Research Scientist Evaluating Variation in Language (EVL) Lab
Eric Shupps is now defaming and libeling me on the Internet along with Amanda Lollar of Bat World Sanctuary. I'll be going downtown tomorrow morning.
