Bug: Word study missing results

1. Navigate to Romans 14.21

2. Select the word meat.

3. Right click on the word choose the lemma and search the entire library. I choose just the Persus Classic collection.

4. I get 919 results in 373 articles in 144 resources.

5. Now run a word study on the same word and look at the results of the classics section under textual searches. I get two. Is this correct?

Using the latest release candidate. No I am not indexing.

Find more posts tagged with

Comments

Graham Criddle

Hi Lynden

I can reproduce this and agree it doesn't seem right.

Graham

MJ. Smith

https://community.logos.com/discussion/comment/455906#Comment_455906

The BWS runs a morphological search; the right click runs an all text match all forms search.

Lynden O. Williams

https://community.logos.com/discussion/comment/456076#Comment_456076

The BWS runs a morphological search; the right click runs an all text match all forms search.

Thanks MJ. However, if I run a morphological search on the word (see screen shot) in the Classics, how is it that I get so many hits?

JT (alabama24)

https://community.logos.com/discussion/comment/456208#Comment_456208

Pardon my ignorance, Lynden… Are these bible verses in the Perseus collection? I wouldn't expect there to be any results in the ESV if they weren't.

Graham Criddle

https://community.logos.com/discussion/comment/456220#Comment_456220

Pardon my ignorance, Lynden… Are these bible verses in the Perseus collection? I wouldn't expect there to be any results in the ESV if they weren't.

I don't think Lynden's point is that there are no hits in the ESV (I think that is just how he has his search environment set up) - it is more that the Classic search from BWS doesn't find more.

Graham Criddle

https://community.logos.com/discussion/comment/456208#Comment_456208

Looking at the log file after trying this search I see:

013-01-27 14:21:00.0202 1 Info PanelViewModel Navigating panel to: BibleReferenceRange=|ExpandedStates=ClearDefaults%3dtrue|MatchAllWordForms=true|MatchCase=false|
MorphReferenceRange=|MorphResourceProvider=Temporary%7cResourceIds%3dLLS:EURBAGK,LLS:PLATLAWSGK%7cTitle%3dClassics|QueryText=<lemma%20%3d%20lbs/el/ÎºÏÎÎ±Ï‚>|SearchKindName=MorphSearch

This seems to suggest that the search is just executing against two resources - EURBAGK (Bacchae (Greek)) and PLATLAWSGK (Laws (Greek))

If this is the case then it explains the results we are seeing.

So the key question is how the "Classics" in BWS is / should be defined

JT (alabama24)

https://community.logos.com/discussion/comment/456241#Comment_456241

I did say, "pardon my ignorance," right? [:D]

(Hey, it's all Greek to me!)

Graham Criddle

https://community.logos.com/discussion/comment/456297#Comment_456297

I did say, "pardon my ignorance," right?

Fair point[:D]

MJ. Smith

https://community.logos.com/discussion/comment/456317#Comment_456317

How much of Perseus do you think it tagged with Logos Greek morphology? If I wasn't feeling lazy, I would put my results in aligned view to see what mix I am actually getting of the lemma vs. inflected forms.

Graham Criddle

https://community.logos.com/discussion/comment/456208#Comment_456208

Hi Lynden

Thanks MJ. However, if I run a morphological search on the word (see screen shot) in the Classics, how is it that I get so many hits?

Just worth noting that the search you have here is against "Perseus Classics" whereas the search generated from the BWS is for "Classics"

As per my earlier note I think this is the root of the problem.

Graham

Graham Criddle

https://community.logos.com/discussion/comment/456372#Comment_456372

How much of Perseus do you think it tagged with Logos Greek morphology?

From the results we are seeing - I think, quite a few. According to the blog post announcing Perseus (http://blog.logos.com/2011/08/3000-free-perseus-books-for-logos-4/) Logos were going to include "lemma tagging" and maybe "morph tagging". Not sure if this distinction affects anything here!

The top result in Lynden's search above is from "Acharnians" and I have checked this is tagged.

I would put my results in aligned view to see what mix I am actually getting of the lemma vs. inflected forms.

Doing this shows a variety of inflected forms being returned suggesting that it is correctly searching against the lemma.

As far as I can see the issue is still related to what resources are included in the "Classics" link on the BWS - but I could easily be wrong[:)]

MJ. Smith

https://community.logos.com/discussion/comment/456487#Comment_456487

As far as I can see the issue is still related to what resources are included in the "Classics" link on the BWS - but I could easily be wrong

I suspect you are correct - we've pretty much eliminated the other likely culprits.

Bradley Grainger (Logos)

3. Right click on the word choose the lemma and search the entire library. I choose just the Persus Classic collection.

4. I get 919 results in 373 articles in 144 resources.

5. Now run a word study on the same word and look at the results of the classics section under textual searches. I get two. Is this correct?

I can reproduce this and will have someone look into the underlying problem.

Graham Criddle

https://community.logos.com/discussion/comment/456605#Comment_456605

3. Right click on the word choose the lemma and search the entire library. I choose just the Persus Classic collection.

4. I get 919 results in 373 articles in 144 resources.

5. Now run a word study on the same word and look at the results of the classics section under textual searches. I get two. Is this correct?

I can reproduce this and will have someone look into the underlying problem.

Thanks Bradley

I think we are quite keen to see what is going on here!

Graham

Mark North

https://community.logos.com/discussion/comment/456487#Comment_456487

This brings up a connected point. How do we know ahead of time, when choosing our add-on books, how heavily they are tagged and integrated into Logos? There should be some kind of tags, or a rating of 1-5 to convey that to the consumer.

Dave Hooton

https://community.logos.com/discussion/comment/457359#Comment_457359

That is a good suggestion - probably needs to be put in the Suggestions forum as well.