Author: Jaimie Murdock

  • Reflections on Privacy

    For many people, the primary privacy concern is the "no parents" concept – we don’t care who sees things as long as our "parents" don’t see it (where parents can be anyone we don’t want to see things – professional contacts, straight-edge acquaintences, terrorists, Julian Assange, etc.). This is what I term the exclusive privacy model: start with the public and begin cutting people out. However, this "public minus parents" idea doesn’t make sense. Online, you just have to logout to see this information. Offline, all someone has to do is talk. Facebook was originally marketed this way: here is a place to post information where only Harvard/Ivy League/college students can see it.

    This exclusive model is the most common privacy misperception. Information spreads, and by consciously recognizing this privacy becomes synonymous with trust. For example, you send an e-mail, confide in a friend, or upload a photo. This is private information, but is capable of being shared or forwarded in any number of ways, both online and offline (e.g., gossip). Its reach is mitigated by social convention and our own discretion.

    Google+ gets this inclusive privacy model right. First, it always explicitly states who an item is being shared with, not who is being excluded. When resharing an item that was shared with a limited circle, it notifies you of the original intent, highlighting the priviledge and trust placed in you. Just like an e-mail program’s forward button, each piece of content has a share button and the API will allow for all data to be federated outside of Google+. However, you also can disable the reshare for each posting. Someone else can always copy-paste your content, but it won’t be computationally linked to you.

    Privacy isn’t just about information, it’s about image as well. Google+ enables full control over your profile. Instead of posting to your wall or tagging you in a photo, people communicate with you directly through limited shares which do not appear on your public profile. Photo tags don’t appear in your albums until they are approved. A box in the upper right corner of your profiles allows you to view it as any other user. Voyeurism is all but eliminated, as you do not see a constant stream of external interactions. Facebook has some of these settings, but they are not as pervasive in the profile.

    The Next Step

    Google+ seems to have figured out a better way to handle privacy – both in terms of information and image – but the next social networking revolution is targeting: I don’t care who sees what I post, but I am self-conscious about overloading people with irrelevant information. My ideal publishing model wouldn’t be about circles of people, but streams of tagged content. If there existed a service where you could follow a person, but mute certain content streams (such as local events, politics, etc.), we’d have perfection. For example, friends in Kentucky don’t care about tornadoes around Bloomington. Professional contacts may be extremely interested in my philosophy and technology content, but don’t care about what concerts I’m going to. People who aren’t in the same circles (hometown friends, college friends, professional contacts, etc.) may share interests in internet humor or politics, while others consider unfollowing me because of it. None of this information is private, but I don’t want to innundate the world with extraneous chatter. If a social network can figure this out, that’s where I’ll plant my flag.

  • Speciation and Information Theory

    For the past two semesters, I’ve been doing some exploratory work marrying speciation with information theory in the framework of the Polyworld artificial life simulator. The simulation gives us a nice framework for mathematically “pure” evolutionary theory and exploration of neural complexity. We’ve applied clustering algorithms to the genetic information, revealing evidence of both sympatric and allopatric speciation events. The key algorithmic intuition is that genes which are highly selected for will conserve, while those which are not will descend to a random distribution (and thus high entropy), so each dimension (gene) can be weighted by its information certainty to alleviate the curse of dimensionality.

    The work was accepted as a poster and extended abstract for the Genetic and Evolutionary Computing Conference (GECCO), and was accepted as a full paper for the European Conference on Artificial Life (ECAL). The full paper is substantially revised from the initial GECCO submission, and provides an introduction to several problems of biological, computational, and information theoretic importance. The visualizations, including several videos showing the cluster data, were especially fun to create, and I’m proud of the finished product.

    There are still several more research directions from this work: the allopatric and sympatric effects have not been differentiated, only one environment was analyzed (consistent with past work on evolution of complexity), the clustering algorithm’s thresholds were not explored for hierarchical effects, alternate clustering algorithms were not explored (future open-source project for me: clusterlib), … Still, the present work is encapsuled, the source is in the Polyworld trunk, and it was accepted for publication.

    Abstract, citation, and paper follow.

    Complex artificial life simulations can yield substantially distinct populations of agents corresponding to different adaptations to a common environment or specialized adaptations to different environments. Here we show how a standard clustering algorithm applied to the artificial genomes of such agents can be used to discover and characterize these subpopulations. As gene changes propagate throughout the population, new subpopulations are produced, which show up as new clusters. Cluster centroids allow us to characterize these different subpopulations and identify their distinct adaptation mechanisms. We suggest these subpopulations may reasonably be thought of as species, even if the simulation software allows interbreeding between members of the different subpopulations, and provide evidence of both sympatric and allopatric speciation in the Polyworld artificial life system. Analyzing intra- and inter-cluster fecundity differences and offspring production rates suggests that speciation is being promoted by a combination of post-zygotic selection (lower fitness of hybrid offspring) and pre-zygotic selection (assortative mating), which may be fostered by reinforcement (the Wallace effect).

    Jaimie Murdock and Larry Yaeger. Identifying Species by Genetic Clustering. In Proceedings of the 2011 European Conference on Artificial Life. Paris, France, 2011. [paper]

  • Spring 2011

    With the passing of another semester comes another life update post. Even though I am no longer a student, being embedded in academia means progress is still measured by semesters.

    Recently, I was awarded the Provost’s Award for Undergraduate Research and Creative Activity, which was a really nice capstone on my undergraduate experience. Since I did not walk at graduation, the Honors Convocation was a good opportunity to give my family closure on this chapter of my life.

    Throughout these few months, I’ve been busy writing up a storm – one week in April saw 30 pages of manuscripts submitted. My previous post details the accepted poster summary on "Genetic Clustering for Species Identification" and the accepted book chapter on "Evaluating Dynamic Ontologies". There are two more papers in review and preparation right now. One is an expansion of the speciation work for a (hopeful) full-paper presentation. The other details work on taxonomy alignment carried out this semester.

    I’ve still been travelling a ton. In December, I headed to Berkeley for my first California Christmas with Dad and Justin and my first non-business trip in 4 months. Three weeks later, I went back to California for a site visit at the Stanford Encyclopedia of Philosophy, Big Data Camp, and the O’Reilly Strata Conference. Strata was amazing – learned a ton, and met some really great people. Definitely planning to go again next year. I was scheduled to go to the Digital Humanities API Workshop but snow delays forced me to cancel, and last minute logistics chagnes made PyCon and ThatCamp SE impossible to attend. These three were certainly disappointments, but after being in an airport every month for 8 months, it was kind of nice to stay rooted for a while. Earlier this week, I visited Princeton University and Beth Mardutho: The Syriac Institute, as part of my work with the Syriac Reference Portal.

    On a more personal note, the diaspora of friends has been steadily widening since graduation, including my roommate of 3 years. This has been disturbed, however, by just as many friends changing their plans to either stay in Bloomington or move back. While we will no longer have a single house to hang out in all the time, I’m excited about the social continuity next year.

  • Two New Publications

    This past week brought two publication deadlines, a conference submission deadline, and preparation for a software demo at Harvard. Needless to say, I am exhausted, but it was well worth the effort.

    The first publication is a 2-page summary of work I’ve been doing with Prof. Larry Yaeger looking at speciation mechanisms in artificial life simulations. This was a condesnation of a paper submission for the Genetic and Evolutionary Computing Conference, and I’m really pleased with how much we were able to squeeze in. Abstract, citation, and link follow:

    Artificial life simulations can yield distinct populations of agents representing different adaptations to a common environment or specialized adaptations to different environments. Here we apply a standard clustering algorithm to the genomes of such agents to discover and characterize these subpopulations. As evolution proceeds new subpopulations are produced, which show up as new clusters. Cluster centroids allow us to characterize these different subpopulations and identify their distinct adaptation mechanisms. We suggest these subpopulations may reasonably be thought of as species, even if the simulation software allows interbreeding between members of the different subpopulations. Our results indicate both sympatric and allopatric speciation are present in the Polyworld artificial life system. Our analysis suggests that intra- and inter-cluster fecundity differences may be sufficient to foster sympatric speciation in artificial and biological ecosystems.

    Jaimie Murdock and Larry Yaeger. Genetic Clustering for Species Identification. In Proceedings of the Genetic and Ecolutionary Computation Conference (GECCO) 2011. Dublin, Ireland, 2011. [paper]

    The second publication is an expansion of the work on ontology evaluation presented last year at the 2010 International Conference on Knowledge Engineering and Ontology Development (KEOD) in Valencia, Spain. We’ve completely rewritten the section on our volatility score, and tightened up the language throughout. The 20-page behemoth will be published as a chapter in an upcoming volume of Springer-Verlag’s Communications in Computer and Information Science (CCIS) series. Abstract, citation, and link follow:

    Ontology evaluation poses a number of difficult challenges requiring different evaluation methodologies, particularly for a "dynamic ontology" generated by a combination of automatic and semi-automatic methods. We review evaluation methods that focus solely on syntactic (formal) correctness, on the preservation of semantic structure, or on pragmatic utility. We propose two novel methods for dynamic ontology evaluation and describe the use of these methods for evaluating the different taxonomic representations that are generated at different times or with different amounts of expert feedback. These methods are then applied to the Indiana Philosophy Ontology (InPhO), and used to guide the ontology enrichment process.

    Jaimie Murdock, Cameron Buckner and Colin Allen. Evaluating Dynamic Ontologies. Communications in Computer and Information Science (Lecture Notes). Spencer-Verlag. 2011. [chapter]

  • Farewell to a Friend

    Today is the memorial service for Helga Keller, a dear friend who has changed so many lives. Since the first weekend in Bloomington, Helga has been a surrogate grandmother for me. We met after church the first weekend I was here and had an instant bond: A German immigrant, her first home in America was my hometown – Murray, Kentucky – and she knew many friends from home. She also was the administrative assistant for Douglas Hofstadter, one of the major inspirations for coming to Indiana. These common bonds of faith, people, and place brought us together throughout the years.

    There are so many things for which I am extremely grateful to her. One day I sent her an e-mail inquiring about the CopyCat program and where I could find articles about it. She responded with a three-page e-mail, with the article attached, links  to all subsequent research, contact information for all of the authors, an offer to introduce me to them, and an invitation to the CRCC lab meetings. As if that weren’t enough, the next time I encountered her she gave me an autographed copy of the book the study appeared in, along with photocopies of the articles mentioned in the e-mail.

    This is but one of many stories of her overwhelming kindness and dedication. May she rest in peace.

  • 2010 in Music

    A year ago, I reflected upon my musical evolution during 2009. As another year passes us by, it’s time to dive into my last.fm data once again for an empirical reality-check on what I’m listening to:

    ‘10 Artist ‘09 Change
    1 The Avett Brothers 3 (+2)
    2 John Mayer 29 (+27)
    3 Radiohead 2 (-1)
    4 The Black Keys 40 (+36)
    5 The Beatles 5
    6 Laura Veirs 52 (+46)
    7 Wilco 1 (-6)
    8 Say Hi 6 (-2)
    9 Kings of Leon 22 (+13)
    10 Nada Surf 10

    All in all, this year was a transition year for music, and I feel it’s continuing to move. Whereas 2009 was largely a solidification of 2008’s favorites, 2010 saw dramatic shifts, with 4 bands making double digit jumps, and 5 bands falling off the top 10 entirely. The only music from last year’s top 10 which became more popular was The Avett Brothers, who leaped to a commanding first place lead on the strength of I and Love and You and a belated discovery of Four Theives Gone.

    As a whole, this year was a move towards a stronger rock influence, as opposed to a folk influence. John Mayer, The Black Keys, Kings of Leon, and the Gaslight Anthem have all landed in my favorite artist lists. In particular, three albums have stood out as essential listening: The Black Keys’ Brothers, The Gaslight Anthem’s The ’59 Sound, and Kings of Leon’s Because of the Times.

    This year I turned 21, which opened up some new opportunities for concerts. Electric Six, Regina Spector, Guster, The Avett Brothers, John Mayer, The Tallest Man on Earth, Joe Pug, and Sufjan Stevens were all particularly memorable nights. Sufjan’s concert was mind-blowingly amazing, both in terms of on-stage production and overall musicianship. I had not heard Age of Adz until that night, but the epic 25-minute “Impossible Soul” is one of the coolest experiments ever. (Sufjan ranked 12th in 2010.)

  • Gentoo: Subversion not permanently accepting SSL certs

    Today I had a rather frustrating issue, as svn would not allow me to permanently accept an SSL Cert under Gentoo, rather just offering me the option to reject or accept temporarily.

    Error validating server certificate for \'xxxxxxxx\':
     - The certificate is not issued by a trusted authority. Use the
       fingerprint to validate the certificate manually!
     - The certificate has an unknown error.
    Certificate information:
     - Hostname: xxxxxxxx
     - Valid: from xxxxxxxx until xxxxxxxx
     - Issuer: xxxxxxxx
     - Fingerprint: xxxxxxxx
    (R)eject or accept (t)emporarily?
    

    After some Googling, I found Bug 295617: subversion won’t save bad certificates permanently with Neon 0.29. By this point Neon 0.28 had left the portage tree, so downgrading was not an easy option. However, a comment on Bug 238529 hinted at a workaround: build Neon without GnuTLS.

    To fix this issue, the easy fix is:

    echo \'net-libs/neon -gnutls\' >> /etc/portage/package.use
    emerge -DN subversion
    

    Neon should rebuild and all will be well!

    Error validating server certificate for \'xxxxxxxx\':
     - The certificate is not issued by a trusted authority. Use the
       fingerprint to validate the certificate manually!
     - The certificate has an unknown error.
    Certificate information:
     - Hostname: xxxxxxxx
     - Valid: from xxxxxxxx until xxxxxxxx
     - Issuer: xxxxxxxx
     - Fingerprint: xxxxxxxx
    (R)eject, accept (t)emporarily or accept (p)ermanently?
    
  • Graduation

    Final grades are in, so I can finally announce that on December 17, 2010, I graduated from Indiana University with dual degrees and honors in Cognitive Science and Computer Science after 7 semesters.

    I’m extraordinarilly excited to finally be done with coursework so I can focus entirely on research. I’ll be continuing work with Prof. Colin Allen and the Indiana Philosophy Ontology Project (InPhO), completing our integration with the Stanford Encyclpedia of Philosophy (SEP) and working on further refinements of the dynamic ontology methodology, generalizing our methods for use in other disciplines. Starting in January, I’ll be working with Prof. David Michelson at the University of Alabama to redeploy the InPhO for the Syriac Reference Portal (SRP). This will hopefully lead to extended collaborations in the digital humanities.

    Additionally, I’m planning to continue contributing to Prof. Larry Yaeger‘s Polyworld Project, which is an Artificial Life simulation that provides a framework for replicable studies in evolution, genetics, and neural networks. I’ve been working on methods of species identification, using information-theoretic measures of genetic distance. This has led to a series of complexity improvements to a popular clustering algorithm used in bio-informatics. I’ve also built a data-access library in Python to facilitate analysis and visualization of experimental data.

    Sometime last year I started travelling all the time. My work was presented in Valencia, Evansville, and Chicago, and I further went to DC, Berkeley, Palo Alto, Louisville, Nashville, and Madrid. So far I’ve got three big trips planned for 2011: Santa Clara in January for the O’Reilly Strata Conference, DC and Philadelphia in February, and Atlanta in March for PyCon. Over the summer, I’ll hopefully be headed to conferences in Ireland and San Fransisco, but we’ll see how that goes.

    Past that, my future plans are predicated on the results of my Fulbright proposal. If this comes through, I’ll head to Karlsruhe, Germany in July to spend a year as a research assistant, developing methods for ontology-driven machine translation and sentiment analysis in collaboratively-generated corpora. Either way, 2011 should be a great year!

  • Published!

    In June, my paper "Two Methods for Evaluating Dynamic Ontologies" was accepted to the 2nd Knowledge Engineering and Ontology Development (KEOD) Conference in Valencia, Spain on October 25-28. The paper was co-authored with Cameron Buckner, a graduate student in Philosophy, and Colin Allen, a Professor in Cognitive Science and History & Philosophy of Science, and details some of our work with the Indiana Philosophy Ontology (InPhO) Project.

    This paper is the culmination of two summers of research on knowledge representation. If you’re interested in the InPhO project, section 3 of the paper is a reasonably accessible summary. The paper as a whole deals with a subproblem in ontologies – how do you quantify the quality of a candidate knowledge representation? We hypothesize that the structure of a domain corpus should be reflected in the structure of a taxonomy of that domain, and that a better taxonomy will better match the corpus statistics.

    I’ll be headed to Valencia October 22-31, and the Hutton Honors College has generously approved a travel grant to cover expenses for the week. I’ve set up my flights to and from Madrid, and I’ll have 2 days before and 3 days after the conference to wander around Spain — I’ve never been to Europe before, so I’m extremely excited!

    The abstract is below:

    Ontology evaluation poses a number of difficult challenges requiring different evaluation methodologies, particularly for a "dynamic ontology" representing a complex set of concepts and generated by a combination of automatic and semi-automatic methods. We review evaluation methods that focus solely on syntactic (formal) correctness, on the preservation of semantic structure, or on pragmatic utility. We propose two novel methods for dynamic ontology evaluation and describe the use of these methods for evaluating the different taxonomic representations that are generated at different times or with different amounts of expert feedback. The proposed "volatility" and "violation" scores represent an attempt to merge syntactic and semantic considerations. Volatility calculates the stability of the methods for ontology generation and extension. Violation measures the degree of "ontological fit" to a text corpus representative of the domain. Combined, they support estimation of convergence towards a stable representation of the domain. No method of evaluation can avoid making substantive normative assumptions about what constitutes "correct" representation, but rendering those assumptions explicit can help with the decision about which methods are appropriate for selecting amongst a set of available ontologies or for tuning the design of methods used to generate a hierarchically organized representation of a domain.

  • virtualbox-bin in Gentoo

    Some non-Linux-dork posts are in the pipe, but today I had issues getting VirtualBox up and running on Gentoo. Here’s some proper install instructions to work around Bug 283617. I’ll get to fixing the ebuild later this weekend.

    emerge virtualbox-bin
    chmod 4750 /opt/VirtualBox/VBoxNetAdpCtl
    chmod 4510 /opt/VirtualBox/VBoxSDL /opt/VirtualBox/VBoxHeadless /opt/VirtualBox/VirtualBox
    gpasswd -a youruser vboxusers
    

    After a logout/login, VirtualBox should appear in your Applications menu, and can be run from the command line with VirtualBox.