BUG? Unexpected result from Place lookup

Right-clicked on "Kedar" in Bible text, selected Place Kedar (region) > Bible Facts.
What came up was a picture of an Arab Man on a Camel. The word "Kedar" not mentioned anywhere on the image.
What I wanted/expected to come up by default would be a map of that region. That was the second image thumbnail, so I was able to click on it and get to where I wanted relatively quickly. But come on, can't you do a better job of guessing what the user wants to see when they look up a place name?
Comments
-
Rosie Perera said:
Right-clicked on "Kedar" in Bible text, selected Place Kedar (region) > Bible Facts.
What came up was a picture of an Arab Man on a Camel. The word "Kedar" not mentioned anywhere on the image.
Replicated puzzling result, then found image description has Kedar:
Wish image description would appear in Bible Facts.
Rosie Perera said:What I wanted/expected to come up by default would be a map of that region.
+1 [Y] for map to appear first for a Bible Fact that includes "region".
Keep Smiling [:)]
0 -
Good find. That does explain the puzzling result, somewhat. However the "Kedar" in this particular image description is a person not a place, so it's irrelevant to the lookup I was doing. It might have made sense if I'd been looking up Kedar (son of Ishmael).
But even that is a very tenuous connection, given the textual description, but I don't expect Logos to be smart enough until maybe version 7 to semantically understand that this image of an Arab on a camel really has nothing to do with Kedar the person even though the word "Kedar" happens to appear in the description. Would need a whole complicated level of tagging of images for semantic relationships. (And I wouldn't want to have to learn a whole new geeky syntax to do searches like this. Just be smart and DWIM.)
I totally agree that the image description should show up in Bible Facts. And maps should have priority over other images when it comes to looking up a place (e.g., city, kingdom, region).
0 -
Rosie Perera said:
Good find. That does explain the puzzling result, somewhat. However the "Kedar" in this particular image description is a person not a place, so it's irrelevant to the lookup I was doing. It might have made sense if I'd been looking up Kedar (son of Ishmael).
I can reproduce this bug, and I will create a case for Development.
Rosie Perera said:I totally agree that the image description should show up in Bible Facts. And maps should have priority over other images when it comes to looking up a place (e.g., city, kingdom, region).
I will also create a case for both of these suggestions.
0 -
Actually there is an image description that appears in the Bible Facts. For example lets use the image of "Kedar (region)". When you hover over the hyperlink "Images from the Temple Dictionary of the Bible" it will display the image description.
But I'll still create a case for maps having a priority over other images.
0 -
Angela Murashov said:
But I'll still create a case for maps having a priority over other images.
And another case for the fact that images related to a person with a particular name (Kedar in this case) shouldn't be suggested at all when searching for a place with the same name.
0 -
Rosie Perera said:
And another case for the fact that images related to a person with a particular name (Kedar in this case) shouldn't be suggested at all when searching for a place with the same name.
Do you know of any other searches that might have this issue? I already have created a case about Kedar.
0 -
Angela Murashov said:
Do you know of any other searches that might have this issue? I already have created a case about Kedar.
Here are a couple more potential ones but it turns out that in both of these cases that there are no images related to the person with the same name, so maps do come up first when you search.
Asshur (place); Asshur (person)
Dan (city); Dan (son of Jacob)
I can't think off the top of my head of any other places and people who share names. There are regions named for all the other tribes of Israel too, but these all show up in the Biblical Facts database as names of the people groups who live there rather than places (e.g., Benjaminites, Naphtalites). Esau (Edom) the person and Edom the place are differentiated enough that the issue doesn't come up there.
I would think if the developers solved the Kedar problem with a general solution rather than a specific hack then it would catch any other instances if such should come along.
0 -
Rosie Perera said:
I would think if the developers solved the Kedar problem with a general solution rather than a specific hack then it would catch any other instances if such should come along.
Yes the results are displaying the correct images for these potential searches. So hopefully "Kedar" is the only issue. Thank you for reporting it! [:)]
0 -
This issue has been fixed in the 2014-01-24 updated resource "Images from the Temple Dictionary of the Bible" (LLS:IMGTEMPLE) .Angela Murashov said:Rosie Perera said:Good find. That does explain the puzzling result, somewhat. However the "Kedar" in this particular image description is a person not a place, so it's irrelevant to the lookup I was doing. It might have made sense if I'd been looking up Kedar (son of Ishmael).
I can reproduce this bug, and I will create a case for Development
0 -
Angela Murashov said:
This issue has been fixed in the 2014-01-24 updated resource "Images from the Temple Dictionary of the Bible" (LLS:IMGTEMPLE) .
Thanks. That issue is fixed, but there is still a problem with Kedar in LCV. Select "Kedar (region)" on the right side, and lookup of "Kedar (Person)" is proposed for many if not most of the dictionaries on the left. Granted, some resources don't have two separate entries but rather list two definitions under the same entry, and I can understand those ones being listed. But clearly at least the ABD entry which specifically says "Kedar (Person)" shouldn't be there.
0 -
Rosie:
We're still picking apart some resources that have one entry with two different entities, but I'd agree that some of these links are suspect. We'll take a closer look.
0 -
On the larger questions raised by this example (now that i've read the whole thread):
- We're working on better ways to find (Logos) maps given a name. For now, another strategy is to use Image Search, restricting the type to "Maps" (logos4:Search;kind=Image;q=kedar;match=stem). I don't seem to have the map you do, so i can't check to be sure this works, though.
- Some of us are smart enough to know the difference in an image annotation between Kedar the Person and Kedar the Place [:)], and since we're annotating these manually (!), this kind of disambiguation should normally happen. But in this case, i believe there was an error in the image annotation.
- I think there are also some annotation errors in the LCV alignments for the various dictionary articles: we're working on that.
(I'm sure at some point we collated a list of people and places who share a name, but i don't recall off the top of my head where it is. There are more than just a handful. )
0