Sunday, June 7, 2009

Adding your library catalogue results next to Google?

Despite talk about Google-killers such as , semantic web engines , real-time search and social search and social media search engines (e.g. oneriot , Social Mention,WhosTalkin?) , "Fact engines" like Wolfram Alpha when it comes down to it, the instinctive first response of anyone (including many librarians) when looking for something is to Google it, and this isn't likely to change for years.

Libraries have tried to handle the Google dominance in several ways. In one of my first posts I talked about the use of searchplugins, bookmarklets, custom toolbars (Libx, Conduit) , smart keywords that allow users to do searches of library resources without visiting the library portal.

That is good, but it is still limited. Even with such tools available, and a Google search fails, the user needs to acquire the habit of copy and pasting the search into the search bar, click on a bookmarklet etc. Or more likely they would go with whatever Google gave and do not even bother to compare with what they could find in the library.

The same problem lies with tools that require you to highlight terms, right click and select engines from the context menu. An evolution of these tools would be IE8's web accelerator (library examples?), KallOut, Headup, and Mashlogic . True, all you need to do is highlight a term on the page and it will automatically create a popup bubble on the page with the information (see example below using KallOut)

But you still need effort to highlight and right click.

A better approach?

So what is easier? Simple, display the results next to a Google search automatically!

This idea came to me when playing with the greasemonkey script that shows Twitter search results on top of Google results.

There's also a very similar one for OneRiot here.

I then started to notice that many commercial search engines with the same idea. - search interface that combines Twitter and Google results.

Wolfram Alpha Google - Firefox add-on that displays Wolfram alpha results next to Google

Google Enhancer - Trueknowledge - TrueKnowledge is a search engine similar to Wolfram alpha, this firefox add-on will display any relevant results (if any) next to Google search results.

OneRiot on Google (WebMynd) - Firefox plugin - This uses the WebMynd plugin, more about that later.

How about adding your Library catalogue results?

I really wish the usual library vendors providing our libraries with software could give us something similar. Imagine a federated search solution that provided a plugin that overlays results along side google results whenever one searched, etc. I bet you access to library resources would sky rocket!

Lacking that, is there any simple hack one could do to achieve this?

I investigated the possibility of doing the same for Library catalogue results (subscribed electronic resources would be another really interesting idea).

In the end I saw three possible options.

Using Opensearch

It is possible to use something like A9's OpenSearch Client , to create column of results for each opensearch search. Or at least it was in the past. I can imagine hacking up something similar with basic javascript, and output the results in frames.

Using Greasemonkey + JSON Output

Another idea would be to adapt the greasemonkey script that shows Twitter search results on top of Google results. For that to work though, your catalogue should be able to output results in JSON. Unfortunately not many catalogues do this yet, I found one but unfortunately lost the url.

Many library catalogues *do* output search results in RSS. A very ugly hack, I could imagine that might work would be to use a RSS to JSON REST service to convert it to JSON and use that.

Using WebMynd

One could also follow OneRiot's lead and use WebMynd. This is a plugin (Internet explorer and Firefox) that adds results from various sources such as Wikipedia, Scridb, Twitter, Google books etc next to Google and Yahoo results (see below).

WebMynd also includes a optional recorder function that saves every page you visit (screenshot and/or htm) allowing you to search through them.

Out of curiosity, I contacted them and inquired about the possibility of adding library catalogue results. Apparently they can use Data scrapping methods to get the results, but only for a fee . If you can guarantee a certain number of installs in systems you control, you can possibly get it for less or free.

I did notice one quick and dirty way that find work. WebMynd, shows results from "My Top sites".

These are the most common domains that WebMynd recorder saves while you are browsing. WebMynd will then displays results from those domains.

A obvious idea would be this. Install WebMynd, turn on recording and visit only your library opac domain. Then you turn off the recording, so your library domain is a "Top site".

Then when you do a search, results from that "top site" will appear. Unfortunately, I could get it to work only for WorldCat results (see below).

The reason seems to be this, the results are drawn from sites that have being indexed by Yahoo. Unfortunately most library catalogue results are not indexed by common search engines like Yahoo or Google (so called invisible web), and as such if you try the same trick for any opac site, it comes up with nil.

Another drawback is that the Worldcat results are also not "real time" results, they are whatever was in the index the last-time Yahoo spidered the site.

My technical skills are limited, I would love to see if someone could get the ideas here to work, or come up with better ideas.
blog comments powered by Disqus

Share this!

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
Related Posts Plugin for WordPress, Blogger...