HI,
Does IDOL pulls in any metadata by default?
for example, IDOL summary, Title, subject....
Or these fields populated by MB without help of IDOL.
Please let me the know the details.
The information for an asset in IDOL is populated by MediaBin. By default, IDOL summary is derived by IDOL based on the DREContent field supplied by MediaBin. How the IDOL Summary is created can be tweaked by editing the AutonomyIDOLServer.cfg file. You can refer to IDOL documentation for details.
Thank you Todd for reply.
It means only IDOL summary is populated by IDOL and remanining metadata are populated by IDOL.
Am i correct?
I think you meant, remaining metadata is populated by MediaBin. Yes.
Can we tweak IDOL config to modify the default metadata of IDOL Summary?,.
For Ex.. instead of paragraphs, we would like to have keywords extracted, like keyword1, keyword2 etc.,?
I do not understand the use case. I think you are trying to re-use functionality from IDOL for some purpose that it was not intended to perform.
You can only change the data fields from which IDOL will construct a summary. It is not able to detect what you believe are keywords.
If you supply Keyword metadata in MediaBin, then they are indexed into IDOL for searching.
Ok, thank you
If you describe the specific use case in more detail, we may be able to explain the best way to achieve the result using MediaBin.
Thanks Todd.
Here is the use case which was implemented in 4.6.2 and would like to carry forward in 8.1.
in 4.6.2 MB extracts keyowords from document automatically using MetaTagger task.
Keywords are limtited to 10 and comma seperated. This MetaData along with document is pulled by TeamSite from MediaBin and then further sent to web. Search in our website is based out of these keywords.
Since Metatagger is not compatible with MB 8.1 we were trying to implement this use case in another way ie using IDOL so that there is no impact to website functionality.
Hope, this is clear.
The concept of only searching for Kewords is a bit dated. With IDOL it can query all indexed data. So why limit users to only search keywords? The IDOL capability to find what you are looking for based on what you use for search criteria is much more powerful and capable that the previous solution. It will find those assets that have your keywords along with other conceptually similar assets.
So I do not understand why this keyword "extraction" is necessary when users can enter a search that IDOL will perform.
