About

SCIgen is a program that generates random Computer Science research papers, including graphs, figures, and citations. It uses a hand-written context-free grammar to form all elements of the papers. Our aim here is to maximize amusement, rather than coherence.

One useful purpose for such a program is to auto-generate submissions to conferences that you suspect might have very low submission standards. A prime example, which you may recognize from spam in your inbox, is SCI/IIIS and its dozens of co-located conferences (check out the very broad conference description on the WMSCI 2005 website). There's also a list of known bogus conferences. Using SCIgen to generate submissions for conferences like this gives us pleasure to no end. In fact, one of our papers was accepted to SCI 2005! See Examples for more details.

We went to WMSCI 2005. Check out the talks and video. You can find more details in our blog.

Also, check out our 10th anniversary celebration project: SCIpher!

Generate a Random Paper

Want to generate a random CS paper of your own? Type in some optional author names below, and click "Generate".

Author 1:
Author 2:
Author 3:
Author 4:
Author 5:

SCIgen currently supports Latin-1 characters, but not the full Unicode character set.

Examples

Here are two papers we submitted to WMSCI 2005:

• Rooter: A Methodology for the Typical Unification of Access Points and Redundancy (PS, PDF)
  Jeremy Stribling, Daniel Aguayo and Maxwell Krohn
  
  This paper was accepted as a "non-reviewed" paper!
  • Acceptance e-mail
  • A strange follow-up email, along with our response
  • Anthony Liekens sent an inquiry to WMSCI about this situation, and received this response, with an amazing letter (PS, PDF) attached. (Also check out Jeff Erickson's in-depth deconstruction of this letter.)
  • With the many generous donations we received, we paid one conference registration fee of $390.
  • Our registration fee was refunded. See above for the next phase of our plan.

  We received many donations to send us to the conference, so that we can give a randomly-generated talk.
  
  

• The Influence of Probabilistic Methodologies on Networking (PS, PDF)
  Thomer M. Gil
  
  

  For some reason, this paper was rejected. We asked for reviews, and got this response.

Talks

Thanks to the generous donations of 165 people, we went to WMSCI 2005 in Orlando and held our own "technical" session in the same hotel. The (randomly-generated) title of the session was The 6th Annual North American Symposium on Methodologies, Theory, and Information. The session included three randomly-generated talks:

• Harnessing Byzantine Fault Tolerance Using Classical Theory
  Dr. Thaddeus Westerson, Institute for Human Understanding (Max)
• Synthesizing Checksums and Lambda Calculus using Jog
  Dr. Mark Zarqawi, American Freedom University (Jeremy)
• On the Study of the Ethernet
  Franz T. Shenkrishnan, PhD, Network Analysis Laboratories (Dan)

As promised, we videotaped the whole thing. You can download the resulting movie, titled Near Science, below. Movie length: 13:15.

• High quality (AVI: 88 MB, RealMedia: 65 MB):
  Download AVI | Download RM
  Bit Torrent AVI
  AVI Mirrors: MIT (MA) | CMU (PA) | Brown (RI)
  RM Mirrors: MIT (MA) | CMU (PA) | Brown (RI)
• Medium quality (AVI: 48 MB, RealMedia: 42MB):
  Download AVI | Download RM
  Bit Torrent AVI
  Coral cache AVI | Coral cache RM
  AVI Mirrors: MIT (MA) | CMU (PA) | Brown (RI)
  RM Mirrors: MIT (MA) | CMU (PA) | Brown (RI)
• Low quality (AVI: 20 MB, RealMedia: 9 MB):
  Download AVI | Download RM
  Bit Torrent AVI
  Coral cache AVI | Coral cache RM
  AVI Mirrors: MIT (MA) | CMU (PA) | Brown (RI)
  RM Mirrors: MIT (MA) | CMU (PA) | Brown (RI)

Trouble playing the AVI? Try downloading a DivX codec for Windows or Mac, or try the open source VideoLAN player.

You can read more about the trip here, and check out some pictures here.

Many thanks to everyone who made this possible, especially Tadd Torborg and family, Open Clipart, the PDOS research group, and of course all the SCIgen donors.

Code

The code for SCIgen is released under GPL, and is now available via github!

https://github.com/strib/scigen

If you are a time-traveler from 2002 and prefer anonymous CVS, here you go:

% cvs -d :pserver:anoncvs@scigen.csail.mit.edu:/var/cvs login 
Logging in to :pserver:anoncvs@scigen.csail.mit.edu:2401/var/cvs
CVS password: _press return_
% cvs -d :pserver:anoncvs@scigen.csail.mit.edu:/var/cvs co -P scigen

We're still working on documentation and making it more user-friendly, but you should be able to figure most of it out from the code. Here's what you need on your computer to run it (we've run it on FreeBSD and GNU/Linux platforms):

• Perl
• LaTeX/BibTeX
• Gnuplot
• GraphViz

If you would like to contribute code to this project (i.e., by helping us expand our context-free grammar with more sentences, nouns, etc.), please contact us with any patches and we'll apply them if they seem reasonable. We hope to set up a better system sometime in the near future.

Running the code. We've been getting a lot of questions about how to run the code. There are quite a few misleading files in the source -- sorry about that. All you need to do to generate a paper is to run make-latex.pl (also look at make-latex.pl --help). You can also use scigen.pl to generate any arbitrary starting target. See scirules.in for most of the grammar rules.

Donations

As indicated above, one of our generated papers got accepted to WMSCI 2005. Our plan was to go there and give a completely randomly-generated talk, delivered entirely with a straight face. However, this is very expensive for grad students such as ourselves. So, we asked visitors to this site to make small donations toward this dream of ours; the response was overwhelming.

Amount of donations: $2401.43 (after PayPal fees)
Number of donations: 165
Amount of time: 72 hours

We used this money to hold our own session at the same hotel as WMSCI 2005.

Related Work

Other papers:

• Another fantastic submission to SCI 2005, by David Mazières and Eddie Kohler
• Alan Sokal's brilliant hoax article (i.e., the Social Text Affair)
• Researchers in Vienna take down the VIDEA conference
• Justin Zobel raises some questions about the validity of SCI

Other generators:

• gzzt.org's list of the best online generators
• The Dada Engine, another tool that generates random text from context-free grammars
• List of text generators from elsewhere.org (on the right)
• Barath Raghavan's Systems Topic Generator
• An essay generator
• SBIR grant proposal generator
• We initially based SCIgen on Chris Coyne's grammar for high school papers; Chris is now making neat pictures with context-free grammars.

Other SCIgen successes:

• Philip Davis got a paper accepted to the Open Information Science Journal.
• Peter Trifonov got a random paper accepted to the GESTS journal.
• Mikhail Gelfand and the Troitsky Variant newspaper published Rooter in Russian in a nationally accredited Russian scientific journal.
• "Herbert Schlangemann" got a SCIgen paper accepted to the IEEE CSSE 2008 conference.
• Students at Sharif University in Iran got a paper accepted by the Journal of Applied Mathematics and Computation.
• Mathias Ulsar got a paper accepted to the IPSI-BG conference.
• Professor Genco Gülan published a paper in the 3rd International Symposium of Interactive Media Design.

People

We are graduate students in the PDOS research group at MIT CSAIL.

• Jeremy Stribling
• Max Krohn
• Dan Aguayo

Contact us at this email address: scigen-dev at the domain pdos.csail.mit.edu