Skip to main content

I have a question about contextualization.

In my setup, I have a functional location with a machine, and under the machine, there's a sub-station. When I contextualize a document related to the sub-station, why isn't it directly linked to the machine level and the functional location? I understand that I manually contextualize it to the specific station, but since the station is part of the machine and the functional location, it should appear there too.

I think there's a missing link in my data modeling between the functional location, machine, and sub-station. What methods can I use to ensure that when I contextualize a document, it shows up at all levels of the hierarchy linked to that machine?

FYI ​@Uzair Wali 


@Mansoor Ahmed The 'indirect’ link is inferred from the asset path, and is not a materialized link by default. We have plans to support these indirect links in the Search experience in the near future.

In the meanwhile, you can always link those resources i.e. documents to all levels on the asset's path to be able to visualize as direct links on all levels.


@Uzair Wali Thank you for your reply.

 

So, are you saying that if I have 100,000 or many more documents, I need to manually link each one to the substation, machine, and functional location? Is there no way to link them just to the substation and have them automatically associate with the asset or functional location, since the substation is part of that asset? If there's no way to link these documents directly to the asset, it would be very time-consuming and not feasible.

Additionally, if I identify that certain documents contain specific keywords, can these documents be automatically linked to the relevant asset and functional location? Is there any way to make this possible?


@Mansoor Ahmed On your first point, as I mentioned earlier, the Search UI functionality to visualize these 'indirectly’ linked files etc. is expected in a future release of the Search UI, without the user having to do any further 'contextualization’. What I proposed was a workaround, short-mid term until that functionality is available, and in no way a recommendation. If you are using the API/SDK, you can basically do a nested query on the document's related asset, to see if the path contains a higher level asset, to see all documents indirectly linked to it.

On your second query, any custom rule that is not a simple entity matching rule on existing attributes will require a customized workflow with custom mapping rules. So yes, it is possible for anyone responsible for the contextualization workflows/pipelines to specify these custom rules as part of their custom contextualization script.


Reply