Custom search

It is nteresting to see the quick take-up of the custom search engine from Google. Here are some examples that have come over my horizon in the last few days:

  • OpenDOAR is a directory of academic open access repositories based at the University of Nottingham. They have implemented a search across the contents of the repositories they list. As they point out, this search is based on whatever Google has indexed of the repository content, which in turn depends on local implementation/configuration details.
  • Peter Murray put up a search of Dan Chudnov's Planet Code4Lib blogs.
  • Bill Drew has created a Google custom search across his collection of wireless resources:
    I have been working on my Google custom search engine for WLANs and Libraries. It is available at http://wirelesslibraries.blogspot.com. I have integrated the LibWireless list into using the gmane port of the list. It also searches Webjunction, PublicIP.Net, WLAN Central, Wireless Libraries blog, and my old Wireless Librarian website. I am working on finding a way to include library hotspots from JiWire and other sources. [Baby Boomer Librarian: Google Custom Search for WLANs and Libraries]
  • And checkout the Law Library Search Engine.
This seems like a pretty significant move by Google which creates real value. It is interesting to look at it in the context of network flow. The other day, I noted Hitwise's analysis of how Wikipedia channeled a signficant amount of traffic to sites which were downstream of it. At the same time, Wikipedia received quite a bit of traffic from Google, upstream of it. How people move around the web is becoming more interesting.

Custom search looks like a good way of gathering more upstream traffic by placing the ability to search in many more places. The flow to Google will be increased by placing relevant searches in the user flow.

Related entries:

Comments: 4

Oct 27, 2006
Anne Knight

Peter Van Dijck has created an information architecture search engine that has been very helpful to this newbie.

Oct 27, 2006
Bill Drew

Thank you for mentioning my experiments with Google CSE. It is a great tool and may be the most useful product yet from Google.

Oct 27, 2006
Scott Matheson

Co-op is an intersting tool, but it would seem that there should be an easy way to pass your search (successful or unsuccessful) back into a bigger set of the Googleverse.


We do this now, sort of, with our OPACs - we can pass searches out of our OPACs into consortial, union or just neighboring catalogs. It would be nice to be able to do a similar thing in WorldCat.org: search just one library's holdings, a group, or the whole database. I know we can do some of this now with Group Services, but there doesn't seem to be an easy way to move between meta-sets of records. That is, if I search in the Alaska Library Network catalog, I can see that there are 17 holdings of "Cornerstone on College Hill" in the state and even that there are 66 in all of WorldCat, but there isn't an obvious way to see who outside of Alaska has it.


This may seem an odd thing to do, but searching a limited set of records, say from law libraries, would let patrons see what experts in a field (or in a geographic region, as with the Alaska example) think are the important resources. Then a patron might want to find a more local copy that is easier to access.


All of which is to say there is real value in an expertly selected subset of records. That value is enhanced by the ability to easily move in and out of (and among) various subsets of data.

Nov 08, 2006
John Kirriemuir

It is indeed a useful tool. We've created a search version that just searches websites in the Outer Hebrides, containing information only about the Outer Hebrides:

http://www.hebridean.info/

Much better for searching on local things e.g. events, controversies, politicians, than the standard Google index.

It's such a simple device it's a shame Google didn't think of, and release it, a few years ago.