BlogU

  • Vertical Search in Higher Ed

    By Joshua Kim December 28, 2009 10:11 pm EST

    Higher education needs a vertical search engine. This would be a great business opportunity, as a site that got higher ed search right would create a valuable platform for advertising. The value of a lecture capture banner ad would increase for anyone searching about lecture capture. Same with pay-per-click keywords. A vertical search in higher education would also keep readers on the site longer.

     

    It is in writing this Technology and Learning blog that I've come to want a higher ed search engine. Yesterday I was trying to figure out how many hours a day college students spent playing video games. The other day I was looking for levels and trends in LMS adoption (Moodle, Blackboard, D2L etc.). In almost every blog post I write, or presentation I give or article that I'm working on, I need to find some fact, number, statistic, or piece of history related to higher education. Usually, the process of searching for the information I need takes longer then the actual writing. Even worse, I'm often unable to find what I'm looking for.

     

    My search process for higher education research looks something like this. First, I search the box at IHE. IHE has a strong amount of content, and all of it is open and therefore linkable. Next I search The Chronicle of Higher Ed's site. The Chronicle often delivers good material, but much of it is behind a pay wall and therefore cannot be linked (or sent around to colleagues). After IHE and Chronicle search I'm usually forced to go to Google - which unfortunately does a pretty poor job of delivering the relevant and specialized results that I need.

     

    How could a higher ed vertical search engine be created?

     

    Idea 1: A site like IHE could increase the number of "Quick Takes" published each day, paying particular attention to include numbers, levels, and trends in each post.

     

    Idea 2: IHE or the Chronicle could start a new content "bucket" explicitly made to be searched (as opposed to part of the daily browse experience). Perhaps the short items could be crowdsourced or outsourced - removing pre-publishing editorial screening (instead eliminating or editing posts the community flags as inaccurate). The focus could be on capturing stats, levels, and trends - as well as linking to and summarizing a wide range of higher ed. relating news, press releases, and research. Nothing behind a pay wall.

     

    Idea 3: A new "chart and table" graphic content bucket could be created. Simple Excel charts and tables could be made against the existing IHE or Chronicle content - with again all charts free and linkable. Each chart / table could contain the referring link, keywords, and a quick summary. The production of these charts and tables could outsourced or crowdsourced. Making simple charts or tables from existing content seems like a good job to source to a low-wage / high-skill region. Let's say some outsourcing company charged $5 a table/chart (is this realistic) - even if 2,000 charts/tables were made the final cost would be cheap. 2,000 tables or charts in the higher ed vertical would form an excellent base for a specialized search database.

     

    Idea 4: Partner with an existing specialized, human curated search engine such as Mahalo to build higher ed. vertical search pages.

     

    Idea 5: Create a Wikipedia equivalent within the higher ed. vertical on your site. Topic entries are all specific to higher education. Let the community build out all the pages - and then index these pages for search.

     

    How would you go about create the content for a vertical higher ed. search engine? Do you see a need for a service such as this? What content would you like to see populating a database to be searched against?

Go to comments (3) »
Advertisement
On the Job Market Together?

Comments on Vertical Search in Higher Ed

  • Search Engine
  • Posted by djbutler on December 29, 2009 at 8:30am EST

  • When I am looking for a trustworthy statistic from a non-obvious source, I use Google's scholar search engine. I've always been able to find exactly what I am looking for often in the literature review of someone else's research study. The tricky part is belonging to a research institute or organization with access to the online journal where that statistic could be buried.

  • library?
  • Posted by terrapin44 , adjunct prof. at northeast public university on December 29, 2009 at 8:45am EST

  • Have you tried any of the databases that Darmouth subscribes to? I bet if you talked to a librarian their they could hook you up with some good databases. I know we have some that search the Chronicle and similar services and works better then searching the sites individually.

  • Vertical Search
  • Posted by Dorai Thodla , Chief Innovation Officer at iMorph, Inc. on December 30, 2009 at 8:15pm EST

  • It is an interesting problem. It occurs even in other areas like programming (searching for code or libraries). None of the search engines really do a good job, because they are looking to be more general purpose.

    Creating a layer of metadata on top of the content may be a good starting point and leveraging open source search technologies may be an answer. We are trying some of that in our InfoPortal effort.

    First we get very clear targeted content (in a particular domain). We then index it and do the usual keyword search. We submit the resulting information streams to an entity extractor (like open calais) to pull out specific entities. We curate this with humans with expertise in the specific space.

    But your needs go a little beyond. I suspect (but not sure) that the results of the previous step can be fed to an engine like Wolfram Alpha.

    Dorai

Print
© Copyright 2011 Inside Higher Ed