Re: Federate vs. aggregate

"Federated search" has the misfortune of being used various ways by various people.

I think that the Federated Search Blog uses "metasearch" for #1 [1].

[1] http://federatedsearchblog.com/category/metasearch/

-Jodi

On 10 Dec 2010, at 21:25, Karen Coyle wrote:

> Marcia, on the call we defined the terms differently than this:
> 
> 1) federated searching = searching across heterogeneous databases or resources (probably involving post-processing of results)
> 
> 2) aggregated searching = bringing multiple sources into a single database for searching (probably involving pre-processing of metadata)
> 
> You can see this in the minutes of the meeting:
> 
> http://www.w3.org/2005/Incubator/lld/minutes/2010/12/09-lld-minutes.html
> 
> This may mean that we need to revisit our definitions?
> 
> kc
> 
> Quoting "ZENG, MARCIA" <mzeng@kent.edu>:
> 
>> Hi, all,
>> Some thoughts about federated search', in the context of the difference with metasearching.  This is based on an earlier study [ref 1] and I am not comparing it with aggregate specifically.
>> 
>> To the user who does not care where the information is (or who packaged it), the search interface may provide no alert to any difference regardless of the search being processed through metasearching or federated searching (based on a distributed or centralized model). However, the search results of the two models can be decidedly different in terms of quantity, coverage, ranking, and relevance.
>> 
>> The major limitation of metadasearching (e.g., Google) is the access to the deep-Web resources (such as sites that limit access to their pages, sites that require registration and login, and dynamic pages which are returned in response to a submitted query or accessed only through a form). This leads to the concerns over metasearching about the accuracy of searches and the burden that remote searches place on target resources.  That is how federated searching coming in with this context. In federated searching, a wealth of information is incorporated into a single repository and is processed prior to the user's search.  The system then searches a local repository that was created earlier from the previously accumulated data of numerous resources. To a user, the search process itself and the interface of searching may be the same as what he/she has used on major search engines: issues a query, and receive search results.  However, in fact, the results are different because of the quantity, coverage, and ranking processing provided by the federated searching services. Many digital libraries, Web portals, and [it is said] the Google Scholar, are examples of those entities that employ federated searching.
>> 
>> [1] Sadeh, Tamar. 2006. Google Scholar Versus Metasearch Systems. High Energy Physics Libraries Webzine [Online], no. 12. http://library.cern.ch/HEPLW/12/papers/1/
>> 
>> Marcia
>> 
>> On 12/10/10 2:49 AM, "Svensson, Lars" <l.svensson@d-nb.de> wrote:
>> 
>> Dear all,
>> 
>> Sorry for missing out in the meeting yesterday. In the minutes I saw the
>> discussion about federate vs. aggregate:
>> 
>>> <monica> one other way to define federate and aggregate is, federate
>>> is send a search term to remote source, and bring back only results,
>>> aggregate is to pool all possible metadata locally then do the
>>> search.
>> 
>> I'd second that definition, perhaps with the addition, that when you
>> federate (search), you send a search term to remote source_s_, and get
>> results back. But perhaps -- as Monica hinted -- this is too library
>> specific...
>> 
>> All the best,
>> 
>> Lars
>> 
>> --
>> Dr. Lars G. Svensson
>> Deutsche Nationalbibliothek / Informationstechnik
>> http://www.d-nb.de/
>> 
>> 
>> 
>> 
>> 
> 
> 
> 
> -- 
> Karen Coyle
> kcoyle@kcoyle.net http://kcoyle.net
> ph: 1-510-540-7596
> m: 1-510-435-8234
> skype: kcoylenet
> 
>