L'informatique est morte. Vive l'informatique!

Posted by Michael Giarlo on March 19, 2007

Is the discipline of computer science on its last legs? Neil McBride, a lecturer in the School of Computing at De Montfort University, advocates for great change in The death of computing. Citing greatly reduced CS enrollment figures in the UK, US, and Australia, and the growing disconnect between industry and academia, McBride argues that computer scientists need to reform the field from within or risk further marginalization and ultimate irrelevance. Though I can sympathize with his desire to revitalize the discipline, and I understand how perception of computer science might have suffered from the dot-com bust, his message amounts to more than mere doomsaying and pointless nostalgia.

If it is true that computer scientists "look to games programming for [their] salvation," there is a great opportunity being missed. There are other options — exciting, pragmatic, and revolutionary options — computer scientists might investigate if they believe a wholesale rededication of their skills is needed for the betterment of their field. (To be sure, some already have begun this great work.) As a former student of information science and computational linguistics, I'm here as an interested observer to say that your skills are needed if we are to accomplish some of our loftiest goals. I humbly submit the following areas that could use your help:

  • Information retrieval: Build smarter, faster algorithms for finding and organizing information. Instead of building a better bubble sort, figure out better ways to access and relate disparate bits of information. The Google guys made a couple bucks at this; why not cite their success, and point at the meteoric rise of Google, as evidence of the continuing and growing sexiness of computer science?
  • Semantic web: Bring your knowledge to bear upon the growing semantic web discussion. If you could think up distributed computing, perhaps the challenge of distributed networks of semantically encoded data is ready for your insight.
  • Natural language processing: Be the Google-killer by being the first to market with a usable natural language search tool. There is much research in NLP, but very little of it seems ready for end-users. Help make the keyword a thing of the past. Computational linguists would love to cultivate interdisciplinary connections with you folks.

Although McBride's article may fade into the background of the very frequent, if strident, cries that CS is dead, I am hopeful that interdisciplinary ties between computer science, information science, library studies, and linguistics will bring about practical innovation, not to mention a renewed sense of relevance and excitement for computer scientists.

Trackbacks

Use this link to trackback from your own site.

Comments

Leave a response

  1. Shaun Thu, 19 Apr 2007 11:11:04 UTC

    Michael,
    Yes, computer science in academia IS greatly removed from much of the software development industry. It's a shame that some of the challenges the industry faces are not being addressed by computer science departments. However, I wonder what skills in particular do computer scientists posess that would improve research and development of the semantic web, natural language searching, and information retrieval? It would be great if some cross-polination of ideas could occur, but in the end, perhaps Business, IT, and Library Science departments should take the lead in those areas of research and work to acquire the skills necessary to take it to the next level.

    By the way, thanks for the "Rails Way" blog link in your 3/21/2007 post. I haven't gotten to any Rails development yet, but it may not be far away! Excellent resource.

    Great titles for your posts too… I particularly like "Will Libraries Smell Like Teen Spirit?"

  2. mjgiarlo Thu, 19 Apr 2007 11:51:31 UTC

    Hey Shaun,

    I suspect that there already is some cross-pollination taking place, based on my limited experience in academia.

    For instance, one of the professors at SCILS who teaches Information Retrieval came out of a Computer Science Ph.D. program, and the Computational Linguistics program at the University of Washington both has strong ties to the Computer Science and Electrical Engineering departments and encourages computer science students to apply. (The theory being that the linguistics department can better teach linguistics to computer science grads and programmers than they can teach programming to linguistics grads.)

    Computer scientists possess a wealth of skills to be applied to these areas, especially the latter two, NLS and IR. Their skills in higher-level mathematics and training in algorithms and data structures are crucial to NLS and IR innovation, in my humble opinion, and probably cannot come from Library Studies programs alone.

    I'm not suggesting that Computer scientists go it alone. But if they are going through an existential crisis, they should know that there are plenty of interesting, lucrative, sexy, bleeding-edge problems out there to dedicate their considerable skills to.

    Glad you're enjoying the Rails Way! I've got voluminous notes from that series. If you're also interested in the dead-simple way that Rails implements the REST architecture — which I am now coming around to — check out this ongoing series of posts[1].

    The best way to learn Ruby and Rails, IMO, are the canonical books from the Pragmatic Programmer series: Programming Ruby, 2nd Ed.[2] (known as the "pickaxe"), and Agile Web Development with Rails, 2nd Ed.[3]. Definitely worth the cashish if you're serious about learning either! They've been indispensable to my learning, and I now have a love affair with Ruby.

    Thanks for writing in, Shaun. :)

    1. http://www.softiesonrails.com/tags/rest
    2. http://www.pragmaticprogrammer.com/titles/ruby/index.html
    3. http://www.pragmaticprogrammer.com/titles/rails2/index.html

  3. Mycado Wed, 03 Sep 2008 17:35:57 UTC

    Computing need to be the left hand of the men.

Comments