Seems a TYPO in a number of places, both "whichthe" and "whichthey", unless its an old form of valid english thats long-since gone!
Can Logos pick the typo from here, or does someone need to search and file them one-by-one?
Seems a TYPO in a number of places, both "whichthe" and "whichthey", unless its an old form of valid english thats long-since gone! Can Logos pick the typo from here, or does someone need to search and file them one-by-one?
Nope, they don't usually pick them up from here. Someone needs to search and report them one-by-one. I've reported them all in Nicene & Post-Nicene Fathers, Schaff's History of the Church, WBC, and Calvin's Commentaries. If you have other hits in your library, please report those ones.
*** MASSIVE TYPOS LIST ***
OK, using the Search dropdown, there is a massive amount of typos that are in the resources I have.
It would be insane for me to attempt to find and report each in turn!!!
Logos needs to run some kind of global spell-check exception reporting over the master index of their complete library as a part of standard quality control.
If I use "The" and each letter, a..z in the search box, I'm getting 10's of incorrect words comming up.
Please logos, pick these up in your master lists and do some overall quality control of this kind of issues.
( For those I have checked, they seem to be mostly, but not all, in "The Nicene and Post-Nicene Fathers" series. Maybe something with those fonts and your OCR software needs a kick in the right spot. )
Example follows:
The info on these books indicates that Logos produced them in 1997. I imagine that OCR software and Logos' quality control has significantly improved since then. I can't imagine fixing the typos in these resources is all that high on Logos' to-do list...
Pick one in a resource and send the typo with a comment that there are many others - just search.
I can't imagine fixing the typos in these resources is all that high on Logos' to-do list...
Kevin,
I think thats maybe so, but its clear proof that there is no master spell-check/quality-control process on the master word lists generated from the resources.
The problem is two-fold:
A) We get junk in our index dropdown as we type part-words, so it works less well as it only shows the first 20 or so options
A search on a valid word will NEVER find it, if its tacked onto another word.One example is that someone might miss the entry for "Jesus Christ" in the resource "Who's Who in Christian History" when they search on "Christian Church".
(Could I find a better example ?)
The info on these books indicates that Logos produced them in 1997
Also, just for the record, Logos lists the set at US$250 and in fact its one of the deciding factors for the Library package I purchased.
With the typos, what we have in effect is a search that will NOT find an item at times.
These are, of course, scanning errors. One would think the the ETD department would have an automated tool to look for many of these. We need a way, other that reporting many individual typos, to point to classes of things that they need to add to their automated tools.
I wonder how often they rerun the tools to correct new classes of problems in previously published books. Adding 1,000's of new titles is great, but we also need the old ones corrected. If not we will assume that new resources will have the same problems the old one have.
Here is just another example.
*** MASSIVE TYPOS LIST *** OK, using the Search dropdown, there is a massive amount of typos that are in the resources I have
OK, using the Search dropdown, there is a massive amount of typos that are in the resources I have
Logos: Any comments on this aspect of the resources?
We're aware of the ability to use the search engine to find typos across the library. The complicating factor for us is that while the search engine reports one particular typo in dozens of resources, it's much more efficient for us to process dozens of typos in one resource. I hope that we can write some small scripts/programs to check the search engine term list for non-dictionary words, find the instances, and group them by resource for efficient correction. (We do have a team dedicated to fixing existing resources and I will make them aware of this thread if they haven't already seen it.)
I hope that we can write some small scripts/programs to check the search engine term list for non-dictionary words ...
Thanks Bradley.
YES - Thats what I was thinking would be the way to do it. Process the master word lists for each resource, as extracted from the indexing process, and run them against spell-checks. Clearly, you would want to list all kinds of "words" like authors, names, publishers etc.
I think the pay-back in quality of the resources would be well returned for the effort. You already have many of the tools in place (book indexing).
Thanks.
... I hope that we can write some small scripts/programs ...
Sounds like a job for the API in 4.0d [:)]
We're aware of the ability to use the search engine to find typos across the library.
As a related point, the indexing log has the ability to point out tagging errors across the library. Can we assume that ETD are on to this as well?
