[request] Issue with community tagging / Logos Controlled Vocabulary

Okay, I'll admit I regularly complain about the glacial speed (before global warming glacial speed to be precise) of items being added to the LCV. So here's the thing

I would like to be able to search the Talmud for comments by particular rabbis.
I start but discover that the Talmudic rabbis are not yet in the LCV.
But Logos thought of that and gave me the Freebase option (Google type not jump type)
So I tag a number of Rav Papa comments
I foolishly do a search on <Person Rav Papa> and get no results
I should have entered <Freebase Rav Papa [08hxtk]> - not as bad as it looks - the system looks the part in square brackets up
However, when Logos does catch up an add the entry to the LCV, they have not promised a conversion - automatic or user initiated
So now to find everything I need "Rav Papa" OR <Person Rav Papa> OR <Freebase Rav Papa [08hxtk]> in order to get respectively: untagged resources, resources tagged by Logos, resources tagged by myself but not Logos
In short, Faithlife is asking me to build a long term mess for immediate gratification ...

I've also been running into a problem that may be related - the Factbook lacking appropriate synonyms, resulting in unfound entries.

There has to be a way to speed up getting skeleton entries into the LCV so that this mess is not propagated. In addition there needs to be a way for users to request entries and synonyms that results in quick action for items actually causing actual users actual headaches now.

We also need to know how broadly the LCV is intended to spread ... should I expect to see methodological and liturgical terms? theological terms?

Find more posts tagged with

Comments

Dave Hooton

Bump - 1

Eli Evans (Logos)

Much depends on how the system actually gets used (chicken, egg).

We won't leave you high and dry if we start to annex Freebase topics into LCV. If/when that happens, we'll add a record to our reference equivalence dataset, called "DB:UAV" which stands for "Universal Annotation Vocabulary." If we add a "Rav Papa" entry to LCV, then we can mark it equivalent to <Freebase Rav Papa [08hxtk]> and then it won't matter which reference you've added, they'll be treated the same.

I'll alert Sean Boisen to this thread to extend and revise my comments as needed. He'd be the one who would field your specific synonym suggestions.

Sean Boisen

Since we're making admissions, I'll admit I hadn't thought about adding Talmudic rabbis to the LCV before: but now that you've mentioned it, I can see the value. However, I don't expect that to happen soon: more detailed theological terms and other biographies (like Logos authors) will come first. The primary value of the LCV is that it connects together dictionary and other article-sized topical content. So we're likely to prioritize expansion that enables more of that, or that covers terms (primarily conceptual ones) that users search for in the Topic Guide. Longer-term, the LCV also serves to standardize the way we refer to entities, but there are several orders of magnitude more of those than we're currently equipped to tackle (even Freebase isn't "complete" in that regard, and they had tens of millions of dollars in funding before being bought by Google). It's just a hard problem.

I'd also like to have a good way for users to suggest additions to the LCV, though I haven't seen many people (other than you, MJ) requesting such capabilities. So it hasn't made it to the top of the (crowded) list of priorities.

MJ. Smith

https://community.logos.com/discussion/comment/719572#Comment_719572

I agree that it is a difficult problem. I liked the topics.logos.com idea of user updated "also known as" which allowed people to use the terms of their tradition. I also understand why that approach failed - partially it was too far ahead of the users. But if community tagging proves to provide reasonably accurate data (quality not quantity matters) then perhaps user name equations could still work.

How many #1 priorities do you have? I don't think I ever got above 34.[:D] I'm sure the behind the scenes building of datasets is endless. Be assured you are appreciated even though you mainly hear our complaints.

Eli Evans (Logos)

https://community.logos.com/discussion/comment/719575#Comment_719575

More true confessions: We'll probably regret that we chose <Person ...> as the alias for <Entity-From-The-Bible-Knowledgebase-Datasets ...> someday. When that regret comes due, we'll probably change the alias to something more specific. But it should NOT affect the data tagging.

Dave Hooton

https://community.logos.com/discussion/comment/719519#Comment_719519

If/when that happens, we'll add a record to our reference equivalence dataset, called "DB:UAV" which stands for "Universal Annotation Vocabulary."

Thanks for that - I was going to ask sooner or later!

Sean Boisen

https://community.logos.com/discussion/comment/719575#Comment_719575

I agree that it is a difficult problem. I liked the topics.logos.com idea of user updated "also known as" which allowed people to use the terms of their tradition. I also understand why that approach failed - partially it was too far ahead of the users. But if community tagging proves to provide reasonably accurate data (quality not quantity matters) then perhaps user name equations could still work.

How many #1 priorities do you have? I don't think I ever got above 34. I'm sure the behind the scenes building of datasets is endless. Be assured you are appreciated even though you mainly hear our complaints.

I haven't seen any data yet on the usage of Community Tags (investigating that is on my list, but down below 34). The challenge for your use case is the intersection of

how many people use Community Tags?
how much work is it for users to specify the data for a new concept? (just naming a tag isn't enough)
how much additional value does it add? (not much without further annotation and alignment)

Right now, that just makes this a lower priority, though i hope we'll get there eventually.