Bionic Reading implementation on Kobo?

[davidfor]

Originally Posted by PeterT

Actually I have to say that right now I AM reading using OpenDyslexic and enjoying it. With my current eye issues (no vision on right side, trying to limit my contact lens use in left eyes, no prescription glasses (due to kerecatonus in both eyes, hard to fit)) I'm finding OD ideal for me.

When I looked at the page, it didn't appeal to me. But, made me think of OpenDyslexic and how that seems to work for some people. I think I would find this one distracting, but, it appears to works for others. But, honestly, I find it hard to take seriously because I can't take the page seriously. The fact that they have this method of improving readability but don't actually use it, those horrible headlines, and there is no reference to any sort of research, makes it all questionable.

[Uncle Robin]

An interesting thread on Twitter about this

https://twitter.com/jkdempc/status/1527287163796107268

[ohx1b]

I used BR off and on for about ~2 years and I never really drew any conclusions about it. I stopped when I stopped using Reeder and can't really say I missed it. It definitely felt like I was reading faster but also always felt a little bit awkward. I have noticed that when I actively try to read faster I kind of replicate the technique of "latching" on to certain words. I have very mixed feelings lol.

Anyway I actually saw that same tweet and came here specifically to see if anyone had made a KOReader plugin or something. I'm not sure if the api can be used without an internet connection so a Calibre plugin might be a better solution. I don't have any experience with koreader/calibre plugin development though but I might take a crack at it. I'll keep you posted, OP.

[evilone]

Originally Posted by ohx1b

I used BR off and on for about ~2 years and I never really drew any conclusions about it. I stopped when I stopped using Reeder and can't really say I missed it. It definitely felt like I was reading faster but also always felt a little bit awkward. I have noticed that when I actively try to read faster I kind of replicate the technique of "latching" on to certain words. I have very mixed feelings lol.

Anyway I actually saw that same tweet and came here specifically to see if anyone had made a KOReader plugin or something. I'm not sure if the api can be used without an internet connection so a Calibre plugin might be a better solution. I don't have any experience with koreader/calibre plugin development though but I might take a crack at it. I'll keep you posted, OP.

that web maybe give you some helohttps://www.mobileread.com/forums/sh...d.php?t=346917

[ohx1b]

Well, slight update here... In attempt to avoid using the API I spent some time trying to track down someone else trying to replicate the algorithm and everyone on github had just used "highlight the first ~1/2 of each word." Then I checked old versions of the site to see if there was ever a library or some sort of client-side implementation and I managed to secure the last version of the chrome extension before it was taken down (the recent popularity is because of the launch of the paid API so they stopped providing the browser extensions). Anyway long story short I was quite disappointed to learn that the "patented" algorithm is exactly the same as the method mentioned above except the ratio is different and it employs a... specific method of rounding... lol. I find it hard to believe that such an algorithm could actually be patented but to be safe I'll refrain from mentioning exactly what those are for now. Working on a script/plugin at the moment, I just felt like sharing in case anyone else thought the algorithm might be doing something more interesting.

[PeterT]

See https://patents.google.com/patent/DE102017112916A1/en

Sent from my Pixel 4a using Tapatalk

[PeterT]

Also see https://ebookfriendly.com/bionic-reading-kindle-how-to/



Sent from my Pixel 4a using Tapatalk

[iandol]

Originally Posted by ohx1b

Working on a script/plugin at the moment, I just felt like sharing in case anyone else thought the algorithm might be doing something more interesting.

Ha, this is fascinating and depressing at the same time; thanks for sharing this insight!!! I had at least assumed some sort of fixed database of words derived from the use of eye tracking during reading research, but no

@ohx1b - it appears there is a new chrome plugin, or is this the one you looked into: https://chrome.google.com/webstore/d...iecmlflj?hl=de

I am a Neuroscientists who specialises in visual perception, and this follows a long line of tricks to try to improve reading speed or comprehension with variable and conflicting results. BR seems to be developed by a typographer, not a scientist, and the analysis of the underlying algorithm only reinforces this is window dressing more than anything else...

[Quoth]

Originally Posted by jeff47

Just to be clear - I wasn't suggesting it become required for anyone. I find OpenDyslexic very annoying but I don't oppose it being *available*. I've seen (anecdotal) accounts from people with ADHD/ADD who found this very useful, so I was curious.

The fact that you personally don't like it isn't a compelling argument...

I'd not insist OpenDyslexic is removed, but there is no real evidence beyond placebo of any value to it. I could find no proper trials. Anecdotally I tried an ereader with some dyslexic people and they found adjusting the size of Georgia was fine and that OpenDyslexic was harder to read than anything, at any size. They thought eink a big improvement on LCD and being able to set font size and margin vs print book an advantage.

People with no or little training come up with ideas and convince themselves it's world beating.

[Quoth]

Originally Posted by iandol

…and this follows a long line of tricks to try to improve reading speed or comprehension with variable and conflicting results.

Yes, it seems in last 25 years people still arguing about Sans vs Serif. Probably what you are used to reading novels works best? But an unscientific guess.

I always read very fast and never felt any issue of comprehension. Actually learning text to recite it later was a problem. I found when I had to do proof reading for errors I had to learn to read slower as otherwise I'd imagine what ought to be there and miss mistakes:
Missing prepositions.
Missing punctuation.

Missing letters actual type: he, her, place, discus
Intended words: her, here, palace, discuss

Mistyped correct words: think <-> thing etc.

Of course near proof AI is imaginary as Spelling & Grammar checking tools only marginally better than best 30 years ago and proof reading is mandatory. In reality almost all AI is simply pattern matching (not recognition, that needs sentience) usually employing a human curated database. The only thing much changed in 40 years is how the database is built (so-called Machine Learning and Neural Networks, both lying terms because there is no connection to animal learning or real brains)

[JSWolf]

Originally Posted by Quoth

I'd not insist OpenDyslexic is removed, but there is no real evidence beyond placebo of any value to it. I could find no proper trials. Anecdotally I tried an ereader with some dyslexic people and they found adjusting the size of Georgia was fine and that OpenDyslexic was harder to read than anything, at any size. They thought eink a big improvement on LCD and being able to set font size and margin vs print book an advantage.

People with no or little training come up with ideas and convince themselves it's world beating.

https://blog.dyslexia.com/good-fonts...rimental-study

[DNSB]

For me, the following from Good Fonts for Dyslexia – An Experimental Study (the same item as referenced by JSWolf). While the wording may change, the meaning comes through in most the studies I've seen on dyslexia and fonts:

Research Update: A more recent study with dyslexic elementary students confirmed that use of the OpenDyslexic font had no impact on reading ability. There was no improvement in reading rate or accuracy for students using the font, and no students reported preferring the font for reading material.

[JSWolf]

Originally Posted by DNSB

For me, the following from Good Fonts for Dyslexia – An Experimental Study (the same item as referenced by JSWolf). While the wording may change, the meaning comes through in most the studies I've seen on dyslexia and fonts:

The study found that Ariel has the fastest average reading time of the fonts tested and that includes OpenDyslexic.

[Quoth]

Originally Posted by DNSB

For me, the following from Good Fonts for Dyslexia – An Experimental Study (the same item as referenced by JSWolf). While the wording may change, the meaning comes through in most the studies I've seen on dyslexia and fonts:

I'm not surprised.

Also for people mostly reading web sites, Arial style sans fonts are easiest and for people that hardly use web sites or phones, but read lots, the Georgia / Times style serif fonts work best.