(Artificial Intelligence) Tagging/markup all text in all Logos Research Edition resources

I'm not sure of the terminology for this, but say, if you search for "Moses" in 'Everything' you'll get a lot more results from a lot more resources than searching for "<Person Moses>", because it seems certain resources, especially bibles, get much deeper tagging than other resources also classified as "Logos Research Edition." I'm especially desiring of being able to search the library for topics or preaching themes (e.g., seeing which resources cover <Topic Equality> instead of having to search for the word "equality").

I realize that the deep tagging of bible text is only possible because of all the hard work by hand that goes into these resources, so I look forward to AI (I'm sure AI has already been created that can parse language, CleverBot for example) being used to make for much richer tagging (and perhaps cross linking) of all resources, whenever that is possible.

I think I am not understanding how "Logos Research Edition" resources in general are more functional than "Logos Reader Edition" or "E-Book" editions when, if I understand correctly, most resources are mostly plain text only with links to scripture and limited cross referencing to other resources. On that note, I'd like to, say, find who mentions or references the sayings (in a block of text / passage in a resource) of an early church father like Irenaeus in like manner to being able to get a passage guide with detailed information on, say, John 3:16. I don't know of any guides or tools besides basic search that can work with / analyze resources in general besides bibles.

Find more posts tagged with

Comments

Fumbling_Foo

I want to correct myself and say that there is a good deal of guides and tools that can be used with non-bible resources and that "Logos Research Edition" non-bible resources are still pretty rich in markup/tagging. I'm mainly suggesting AI-assisted tagging/markup for search purposes described in the original post above and being able to use the factbook and passage-analysis-like tools with non-bible resources in order to cross reference mentions and commentaries on them. That is, if there's any practical use for this in bible software. The factbook and passage analysis are powerful and awesome tools, and its even possible already to do a "Word Study" or call on the "Topic Guide" in non-bible resources, but I'd like to see these tools for insight into texts brought to bear on most any resource. Also, the search function is already very good at finding relevant resources anyway (but also hopefully there are plans for a general "fuzzy search" in addition to the "fuzzy bible search" ).

MJ. Smith

https://community.logos.com/discussion/comment/1150210#Comment_1150210

Also remember that for links of interest that Faithlife does not make, the Community Tagging options allows the user to add the tags.

Fumbling_Foo

https://community.logos.com/discussion/comment/1150270#Comment_1150270

Thank you so much ! I didn't know about the community tags feature / dataset. I've found a link on this.

https://blog.logos.com/logos-6-tag-your-own-resources/

Replying to my topic here in general, doing a few searches on people and places, it seems that still there are only a few tens out of thousands of non-bible resources in my library that are in the community tags database. Of course these are some of the most useful resources to have community tagging, like dictionaries, but I'd imagine that since Logos 6 the (I reckon large to very large) customer user base has tagged far more than just these resources. Do all community tags everyone does get included in the "Community Tags" dataset, or is there other datasets/databases I can download? I would definitely understand there not being more available given the involved and relatively tedious nature of the task and the sheer amount of resources, text, and references to go through and tag, especially if there's quality checking requirements (e.g. say at least 5 or so out of 'x' user community tags need to agree before a tag gets included in the "Community Tags" dataset). Still, this is a very nice feature as is and I look forward to using it in the future, and I'll be looking to add community tags myself !

Fumbling_Foo

https://community.logos.com/discussion/comment/1150300#Comment_1150300

*a few tens to perhaps hundreds overall out of thousands of non-bible resources in my library for each search query for common people/places like "Jesus" or "Jerusalem", most of these results being from reference works like dictionaries

Fumbling_Foo

https://community.logos.com/discussion/comment/1150301#Comment_1150301

* I should clarify the exact search terms e.g. "<Place ~ Jerusalem>", "<Person Jesus>" , etc.

MJ. Smith

https://community.logos.com/discussion/comment/1150300#Comment_1150300

Community tagging does not include a quality check but I've found very few errors - it's more a matter of different users having different needs so one may tag sparsely, merely ensuring that a block of text will be noted while others are working on a specific text and tag every reference.

J. Remington Bowling

I'd like to, say, find who mentions or references the sayings (in a block of text / passage in a resource) of an early church father like Irenaeus in like manner to being able to get a passage guide with detailed information on, say, John 3:16.

I don't know about the "in like manner" part for how the passage guide structures things, but just in terms of finding x in y, is this not already possible in Logos? If not, I wouldn't think that it requires natural language processing or machine learning to achieve. Seems like it should be possible with just a python script and regex. The need for machine learning might be useful if you also have in mind situations where the author's reference to Irenaeus is more implicit or contextual.

Fumbling_Foo

https://community.logos.com/discussion/comment/1150360#Comment_1150360

Community tagging does not include a quality check but I've found very few errors - it's more a matter of different users having different needs so one may tag sparsely, merely ensuring that a block of text will be noted while others are working on a specific text and tag every reference.

Ok, thanks !

I don't know about the "in like manner" part for how the passage guide structures things, but just in terms of finding x in y, is this not already possible in Logos? If not, I wouldn't think that it requires natural language processing or machine learning to achieve. Seems like it should be possible with just a python script and regex. The need for machine learning might be useful if you also have in mind situations where the author's reference to Irenaeus is more implicit or contextual.

Yes indeed I'm able to find pretty much anything I want to as long as I know the right key words (and probably would get even better results if I learn the advanced search features ! ). I was thinking about the passage guide perhaps being made into a more universal guide that provides linked information on any block of text, but there are indeed other tools already like the factbook and topic guide (and probably many more I'm not aware of, I'm fairly novice with the advanced features of Logos) that provide information on queries from any resource.

Machine learning would definitely be a nice thing and a powerful addition to Logos bible software if it can be implemented, I can only imagine how useful it would be for inter-textual linking among resources for every resource to have tagging metadata for the text (Logos is said to get more powerful with more resources, and with virtually all text in every resource having tagging metadata would add to that even more). so definitely it's a new potential feature I'm looking forward to and definitely what's already possible with Logos without it is very good . I'm sure I'm probably missing out more on what Logos can do because I don't have a lot of experience or training with it than what would be possible with more advanced tagging; the gap in functionality can probably be mostly bridged with what can already be done with enough know-how in Logos' basic search.

MJ. Smith

https://community.logos.com/discussion/comment/1150360#Comment_1150360

just in terms of finding x in y, is this not already possible in Logos? If not, I wouldn't think that it requires natural language processing or machine learning to achieve.

There is a reason that Denise built her system (and I would do so) around basic AI technology - the number of ambiguous terms, the same term being used differently in different traditions, the texts with inconsistent spellings and punctuations . . . language processing without graphs/Bayesian statistics is extremely limited.

J. Remington Bowling

https://community.logos.com/discussion/comment/1150421#Comment_1150421

>>> import spacy

...

done.

Just kidding... But there's no need to reinvent the wheel from scratch. And I think it's a pretty safe bet that FL is already working on NLP (and probably already using it to some degree).

Fumbling_Foo

I just discovered the "Cited By" tool which does what I was looking for with the "Passage Guide" / inter-text linking style functionality. In-text citation notes covers the rest [:)].