Remediating Materials

by Amy Connolly 1 year, 6 months ago

Request for Ideas

I originally posted this in another discussion but it may go unnoticed there; so, per suggestion from Joanna, I'm re-posting here.

Hello Hive-mind, in our District (Bellingham) we are struggling to identify a functional solution for remediating inaccessible text, as well as other common activities (like annotating with voice typing, or digital inking on touch screens) but I'll try to keep this simple to start.

Whether in print or digital, text which cannot be read aloud with universal tools (like Immersive Reader or Read Aloud in Microsoft programs) continues to be a common barrier. As you can probably guess from the above examples, we are a mostly "Microsoft / Windows district", and we are specifically seeking functional solutions for PCs which are used by most students as their 1:1 device.

Although this is a bit tangential to our primary aim of creating accessible, open materials, I'm hoping we can learn from this group and improve our knowedge!

Thanks in advance. 

Read on for additional context for those willing to engage on this topic ...

In other settings, I've used mobile apps (Lens, Claro PDF Pro, etc) and marginally satisfying desktop software (Read & Write Windows' PDF Scanner) as well as browser extensions, such as texthelp's newish Orbit Reader (which may be a progressive web app?). In my experience, the mobile device (iPad) camera is unmatched by the laptops; however, we don't have iPads at secondary levels. I know we could theoretically scan materials on those big multi function devices in the copy room at each school, then remediate these files with software, but the workflow continues to be a challenege... as you can imagine, it's not efficient or pratical in many cases. 

I really appreciated last's weeks emphasis on "functional accessibility", and I'm not giving up on seeking a solution. I know there must be a functional strategy to convert inaccessible text, have it read aloud in practical manner (ideally, while referencing the text in context*), annotate with voice typing or "digital ink", etc... any ideas?

*Side note on text to speech tools Read Aloud and Immersive Reader - a common concern I hear about Immersive Reader is that the user is removed from the source document and unable to reference context. This would lead teachers to promote Read Aloud, which doesn't create an "immersive" view, but this feature is not in OneNote - at least not OneNote Windows 10, which is our districts go-to verson for now... I want to continue workshopping these strategies so that I can recommend the most reliable, functional option for accessing universal text to speech tools.

To bring it back to our Academy, this issue has emerged from an inaccessible ebook which we have adopted. In order to obtain a "more accessible" version the publisher is requiring additional payment. Although it's a bigger discussion, I wonder what role OER could play here... I also wonder about modifications to our tech environment, e.g. should we be making OCR stations (iPad with Lens app in kiosks) in our libraries? 

 

Aujalee Moore 1 year, 6 months ago

I reached out to one of my colleagues and this is what they had to provide: "First, if they are using their copiers to make the scan, they should see if there's an OCR setting. Most modern copiers either have that option, or you can purchase an OCR circuit board. I've checked with pacific office automation, and they have those boards for their copiers. I assume it would be the same for other copier rental companies.

If they open an image PDF in OrbitNote, it will ask if you want it converted. It'll retain the formatting for the most part and do a pretty good job of converting it into accessible text.

Sometimes, you can import the image PDF into your Google Drive. Right-click and select open in Google Docs. It will attempt to do OCR on the image. Results may vary, but I've often used this when I get an IEP from another district that is an image PDF.

TextHelp may still sell read and write gold. There is a PC or Mac version, and it has a pretty robust OCR converter. Once I converted a 200-page document. It took a while, but the document retained the formatting, and you could use text-to-speech on it.

Bookshare is working on a product called pageAI. I've been to a couple of sessions on it. It's not released yet, but you can upload a PDF to it, and it will do an automatic conversion, including math ML and machine-generated alt text. The output will be EPUB. They are using a machine learning model and are still training it. I’ve tried, but they won't let me play with it yet. But I believe it will be free and available to whoever wants to use it.

Hopefully, someone will work with them on their curriculum procurement process, so they get accessible from the start rather than having to pay additional for it.

I hope some of the information above is helpful,"

Amy Connolly 1 year, 6 months ago

Thanks, Aujalee! I will reply in line below...

Your colleague wrote:

"First, if they are using their copiers to make the scan, they should see if there's an OCR setting. Most modern copiers either have that option, or you can purchase an OCR circuit board. I've checked with pacific office automation, and they have those boards for their copiers. I assume it would be the same for other copier rental companies."

That's a great idea. I didn't know it could be an option! I'll try to find out more.

"If they open an image PDF in OrbitNote, it will ask if you want it converted. It'll retain the formatting for the most part and do a pretty good job of converting it into accessible text."

As far as I could tell when testing, OrbitNote only converts documents if you have the paid version. While it is free for educators, it wouldn't be free for all students.

"Sometimes, you can import the image PDF into your Google Drive. Right-click and select open in Google Docs. It will attempt to do OCR on the image. Results may vary, but I've often used this when I get an IEP from another district that is an image PDF."

True that both Google Docs and Word do a decent job of this. Word actually does better, in my option, because it retains more of the formatting while converting to actual text. This is one of the tricks I show teachers to get buy in for working with me! 

Since we're learning about copyright in this Academy, this strategy raises some concerns because we'd could easily be extracting copyrighted material and using in a manner that wasn't intended. Still, I teach this method because it works in a pinch and improves access for those who need it.

"TextHelp may still sell read and write gold. There is a PC or Mac version, and it has a pretty robust OCR converter. Once I converted a 200-page document. It took a while, but the document retained the formatting, and you could use text-to-speech on it."

Texthelp does sell Read & Write for Windows - which was formerly RW Gold, I think. We have a group license, but our implementation has been really challenging. I've seen this tool do a lot of good when it is district-wide, but given our small user group and some of the software bugs, it's been impossible to get it going in a functional manner. And if you wonder why we wouldn't go district-wide to make things easier, it's because of the incompatibilities with tools like OneNote, which our District uses a lot. I am struggling with this dynamic.

"Bookshare is working on a product called pageAI. I've been to a couple of sessions on it. It's not released yet, but you can upload a PDF to it, and it will do an automatic conversion, including math ML and machine-generated alt text. The output will be EPUB. They are using a machine learning model and are still training it. I’ve tried, but they won't let me play with it yet. But I believe it will be free and available to whoever wants to use it."

Neat. I hope to learn more! Bookshare is where I've been focusing energy to address this issue, for those with identified needs, but it still doesn't address the universal access piece... I'll be curious to learn more about pageAI when it becomes available!

"Hopefully, someone will work with them on their curriculum procurement process, so they get accessible from the start rather than having to pay additional for it."

Yes! I hope for that too, haha. I have shared the PALM materials, but it's hard to spread the word - especially when the decision making is distributed rather than centralized. (Charisse, if you're reading this, please take my limited perspective with a grain of salt!)

"I hope some of the information above is helpful,"

Thank you again.

As an aside, I want to give a shout out to Luis and Cynthia for this great, short article on the topic of procuring accessible digital materials and technology. I noted the increased specificity as compared to PALM's "learning materials".

Aujalee Moore 1 year, 6 months ago

Hey, Amy! I checked with another one of my trusted colleagues in Oregon and here is what he had to offer: 

Hi everyone,

I'm still hung up on scanned PDFs, and I'm wondering what the ISBN is for this text. This sounds like something that would be covered by NIMAS. It doesn't sound like interactive web content. It sounds like they may have missed the opportunity to require a NIMAS file set during the acquisition process. NIMAC has said that they're willing to work with publishers on providing NIMAS files when they're not made available to districts. Maybe they can pursue this through NIMAC.

 

Short of getting accessible files from the NIMAC or the publisher, I don't see any way around OCRing the PDFs. I think they'd be better off having staff do this once rather than trying to have students do this individually. Beyond textbook materials, staff should get in the habit of OCRing their materials, either through their office scanner or through software, before sharing their materials. Or they can use better practices for digital materials. Microsoft's Print to PDF feature will give you flat PDFs that aren't accessible. If you use the Adobe PDF generator or export your documents to PDF instead of using the Print to PDF function, you'll be in better shape. Sharing original files is also a good way to go. Word files are generally more accessible than flat PDFs. 

 

Amy mentions having access to other tools in the past, but it sounds like there aren't a lot of options available to her in this situation. I think Bruce shared a lot of really good options. It sounds like OrbitNote is probably closest to having all the features Amy is after. 

 

In terms of OCR options that are generally available to school staff, Adobe Pro is usually my go to. Most school districts have this available on staff computers. Its OCR tool is easy to use and reliably does a decent job. It's not the best, but it's better than a lot of other readily available tools. ABBYY FineReader seems to generally be considered the gold standard for OCR. The desktop versions of Read&Write use a light version of ABBYY for their scanning function, and that also works quite well. 

 

Because OCR is fairly complex and even dedicated OCR tools don't always nail things, I'm generally skeptical about OCR functions built into copy machines. I could be wrong though. It sounds like TTSD has been using this for a while, so Bruce probably has a pretty good feel for the quality of materials produced with this method. 

 

Opening PDFs in Google Docs is really hit and miss. I think it's a good option to keep in your back pocket, but I think teams shouldn't rely on this method as their primary means of OCRing content. 

 

The pageAI tool from Bookshare that Bruce mentioned sounds pretty exciting. I know you're watching the development of this very closely, Bruce, so please let us know if/when you learn about any new developments! 

 

In terms of marking up PDF contents, Microsoft's Edge browser might be a good option for them. You can add text on top of the PDF or use the pen tool to draw or manually markup the document. If the PDF has been OCRd or has accessible text, then Edge will read the content aloud, let you highlight materials, add comments to text, and get definitions for selected words. In general, it seems a lot more functional than working with PDFs inside OneNote (at least OneNote 2016). There isn't a built in function for speech to text, but I'd imagine that the built-in Windows dictation tool could be utilized within the text fields that are available. 

 

That was a longer response than I expected. I got a little lost down the rabbit hole. Hope this helps!

Amy Connolly 1 year, 6 months ago

Hi Aujalee, thank for you continuing to dialogue with me and crowd-source a solution. I appreciate the input from your colleagues! I still have much to learn and respond below to keep checking my understanding / adding context ... hopefully it's helpful to others, too.

We do have a NIMAS source file for this text, yet we are seeking a format that can be accessed with TTS by any student - not just those with qualifying print disabilities. As mentioned in the article linked above, it seems like the increased embrace of UDL is increasing demand for this type of access.

This particular text is currently utilized as an e-book in a proprietary web reader which does not permit functional Read Aloud in Edge. Despite the publisher stating that Read Aloud would work, it appears to be poorly designed - such that it requires several clicks to get to the feature (which is normally just "play"), then it doesn't read in a logical order.

I've since learned (after some persistent questioning) that the publisher became aware of this issue and updated the resource; but they are asking for more money to make that version available, which irks me. But it is the route our District will likely take to improve access for all students.

As an aside, in the name of efficiency, I may blur together multiple requests while seeking the magic solution to meet more than one need. In this case, I was prompted to reach out in response to several simultaneous needs:

  • More universal access to this specific text, for which we have access to NIMAS source file for qualifying students but no universal TTS access

  • Persistent need to educate around ensuring availability of actual text & not flattening PDFs in OneNote, as your colleague described

  • Teachers and OTs frequently asking for a PC equivalent to Snap Type Pro, an iOS app which allows the user to snap a photo of print and annotate using the built in features of the iPad.

Points well taken about the limits of OCR and the need to obtain and share accessible materials from the start. Meanwhile, I know the need will remain despite our best efforts. Regarding remediation options mentioned by your colleague:

  • OrbitNote and Read & Write desktop's OCR Scanner are both good options, if the user has access to the premium features; yet this is only available to staff and select students. Training has been a big barrier, given the limited user base and inconsistencies with the software - or our configuration; but we will continue to work on it.

  • I think Adobe Pro isn't available to most staff due to cost, but I have been wondering about that route...

  • Good to know about ABBYY PDF reader. That's a new one to me and it looks appealing. 

  • I need to get more familiar with Edge's capabilities when it comes to PDFs. This may need the "Snap Type" need if we can get materials scanned...

Debra Fitzgibbons 1 year, 6 months ago

Obviously, purchasing Accessible Educational Materials upfront is preferred. Make sure your agreement states that you want AEM. See the AEM Center webpage for sample language for procuring accessible hard copy and digital text.

How much is an accessible copy from the publisher? Is it less than paying someone to scan and edit it?

Did they look for the ISBN number in Bookshare? If it isn’t there, the NIMAC coordinator can request it be placed there.

Adobe Pro is an option for PDFs.

Don Johnston’s Snap & Read or Texthelp’s PDF Reader may be able to OCR the inaccessible ebook on the fly.

If you open an inaccessible PDF in Orbit Note, it will OCR and save it.

If you are starting with a hard copy of a book or worksheet, a copier with OCR might be the best bet. You can email it to yourself or a student worker to edit. In the community college setting, I sometimes hired the student who needed the book as work study. They would to listen and edit their book before they needed it for class.

Does the school have a printshop? They can make copies quickly.