Which EN-EN dictionaries are morphological? PB740 new firmware

[Mat1]

Hi!

I've just updated my PB740 Ink Pad 3 Pocketbook to 6.8.4473
and I've read that it's now able to use "morphological dictionaries"

morphological dictionaries that will help to bring different forms of words to their basic form. It will significantly simplify their search in dictionaries.

As I've always found laborious and frustrating to edit a plural formal, remove the s in order to get the noun defitinion, I'm really looking forward to it.
But no hint on where I could find any "morphological ditionary".

I've this link to a bunch of dictionaries but are the morphological?
https://e.pcloud.link/publink/show?c...der=4278386439

Could you point me to a English-English morphological dictionary please?

[Moonbase59]

My Pocketbook Era asked me if I wanted the "Morphology for English" installed (can’t remember the exact wording). I said yes, and now it uses the morphology for all English dictionaries I’ve installed, sometimes resulting in real odd hits. ;-)

If it works like Linux (which I don’t know), they might use the Hunspell format and its .aff affinity file for morphology. I seem to have seen some .aff files on my Pocketbook.

EDIT: Yes, if you download the English morphology from Pocketbook, it creates a folder system/morphology/en/ in your device, which contains (hooray!) the Hunspell files:

• index.aff — English morphology rules
• index.dic — English dictionary

A good move, kudos to Pocketbook for using the most-used »standard« Linux dictionary and morphology files! You can (if you can) even update these yourself.

[rkomar]

To nitpick, pocketbook dictionaries have always had support for morphology (it is what the morphems.txt file is for when creating dictionaries). The problem is that none of the dictionaries included anything but rudimentary morphology rules. I've always wondered if the hunspell rules could be translated to morphems.txt format, but never looked into it seriously. If newer firmwares have direct support for the hunspell rules, then I guess that shows that translating them to morphems.txt format isn't practical. Too bad for us who use older pocketbook devices.

[Moonbase59]

You’re correct of course, but I still criticize the decision to make yet another proprietary dictionary format, instead of using either Hunspell or StarDict format. And one that needs Windows tools to convert, too!

Anyway I’m happy they at least use the Hunspell morphology and hyphenation formats, so we can make our own or at least keep them more up-to-date.

A general problem with e-readers is that they typically use very old and outdated dictionaries and hyphenation databases.

[rkomar]

We are getting way off topic, but it is an interesting discussion. I think that dictionaries are like maps, they give you information valid at the time of creation. If you are reading a book written about 150 years ago, then using dictionaries and maps from that time will give you the most accurate information about what is being written about. Using a new concise dictionary for old text, or an old dictionary for new text can be confusing. So, it's probably ideal to have old and new dictionaries available when reading books from a broad range of times.

[Mat1]

I've checked the folders you mentionned:
I have my dictionaries (Webster and Oxford) in system/dictionaries
but I have no system/morphology and no *.aff files and even not morphems.txt

After updating the firmware I haven't been aked anything about dictionary...
Thansk for your help

[Moonbase59]

I might have been mistaken after all in post #2. Could have been "projected hope" and the `.dic` extension just a coincidence but a totally different file format.

Didn’t bother too much anymore since I found out I can have my "standard" StarDict dictionaries I use everyday on the PC on KOReader…

[Mat1]

I've found a way:
Home / Apps / General: Dictionary and click on dictionary icon on top right
then you have a list of dictionary and for each language there is "Download Morphological"
"Morphology has been installed!"

in E:\system\morphology\fr and en:
I've found index.aff and index.dic that you were refering to ealier :-)

As this is generic folder (and not dedicated to one of the dict in E:\system\dictionaries), does that mean that all the dictionary are now morphological?

[eMole]

I guess all your en and fr dictionaries are morpholigical.
And I guess none of the custom dictionaries benefit from it, I don't think they store language information.
But they do store they own morphology info, that's why I prefer them to stardict format in KOReader.

[Mat1]

I'm a bit confused... the "morpho" downloaded is used by which dic?

Until now, I was using the standard ABBYY Lingvo dictionary shipped with the InkPad and the Oxfrod American Dict and both had trouble like finding plural nouns (I had to edit the word, delete the S and enter).

[Moonbase59]

I somehow assume the lookup language decides? Or maybe even the book language.

The PB software usually comes up with really ridiculous suggestions (I have EN and DE morphology installed), independent of the dictionary (i.e., EN-EN or EN-DE). But suggestions are the same for these two dictionaries. So I guess it runs the morphologizer first and then looks up the resulting word(s). Or something…

That was for me the reason to use KOReader for lookups, which finds the words I look for, and not something else entirely. Apart from that, I use the same dictionaries (and lookup sequence) on the PC, so it’s more useful for me.

Morphology is a good thing, but apparently has to be enhanced a lot.

[eMole]

A bit off: I guess morphology works best with heavily conjugated/declinated languages. I use KOReader a lot, OK with EN and DE dictionary lookups, but a mess with RU. Russian I must read with PBReader and a custon dic that includes the correct morphems.txt.