MOVING THIS TO https://github.com/AceCentre/AAC-Corpora-Collecting/
-
Download a demo: https://thinksmartbox.com/product/grid-3/
-
Prediction engine is SwiftKey (you can see the .net sdk files of learned.json)
-
Stores all preferences in C:\Users\Public\Documents\Smartbox\Grid 3\Users\UserName\langCode\Phrases\history.sqlite
UserName - is whatever users have been set by the Grid. Usually there will only be one. LanCode = e.g en-gb
NB: This C:\Users\Public\Documents\ directory is in the Registry - HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Explorer\Shell Folders\Common Document
FOR /F "tokens=3*" %%A IN ('REG.EXE QUERY "HKLM\SOFTWARE\Microsoft\Windows\CurrentVersion\Explorer\Shell Folders" /V "Common Documents" 2^>NUL ^| FIND "REG_SZ"') DO SET CommonDocs=%%B
- Data is in Table PhraseHistory - Phraseid is matched on Table Phrases -id. Each item has a timestamp. So you can do counts on when and how many phrases are in each History. See this SQLlite database for what the History data looks like
- https://uk.tobiidynavox.com/pages/communicator-5-ap
- Download at https://uk.tobiidynavox.com/pages/communicator-5-product-support?tab=1 - eg https://download.mytobiidynavox.com/Communicator/software/5.5.7/TobiiDynavox_CommunicatorSuite_Installer_5.5.7.12303_en-GB.exe
- Settings at C:\Users\userName\AppData\Roaming\Tobii Dynavox\Communicator\5\Users\User 1\Settings\VocabularyUsage
- Cant figure this out. Seems to only update on exiting Communicator.
- Prediction engine is SwiftKey (you can see the .net sdk files of learned.json)
- https://uk.tobiidynavox.com/products/snap-core-first
- A UWP App - So sandboxed - Data can be found at
%LOCALAPPDATA%\Packages\ - uses swiftkey at leaat on windows
- Uses LAM - Language Activity Monitoring.
- PRC defined this format many years ago. More on its structure here: https://aacinstitute.org/language-sample-collection-in-aac/ - https://eric.ed.gov/?id=ED441300
- Its really not designed for this kind of collection. More like every hit. Its going to take quite a lot of parsing - and a lot of knowledge about what buttons are where for it to be useful. For example this is hello world - and being corrected then spoken. (Speak is I think E6).
### CAUTION ###
The following data represents personal communication.
Please respect privacy accordingly.
Automatic Data Logging NuVoice PASS
Version 2.17 2020-09-12
Prentke Romich Company
*[YY-MM-DD=21-05-10]*
11:43:25.017 RECORD ON
11:43:30.697 LOC =I5
11:43:32.217 LOC =G4 [ ]
11:43:32.218 SPE "h"
11:43:32.937 LOC =C3 [ ]
11:43:32.938 SPE "e"
11:43:32.952 LOC =C3 [ ]
11:43:32.953 SPE "e"
11:43:35.657 LOC =J5 [ ]
11:43:36.903 LOC =J4 [ ]
11:43:36.905 SPE "l"
11:43:37.097 LOC =J4 [ ]
11:43:37.098 SPE "l"
11:43:37.577 LOC =I3 [ ]
11:43:37.577 SPE "o"
11:43:38.297 LOC =E6 [ ]
11:43:38.306 SPE " "
11:43:41.737 LOC =B3 [ ]
11:43:41.738 SPE "w"
11:43:41.769 LOC =B3 [ ]
11:43:41.770 SPE "w"
11:43:42.537 LOC =I3 [ ]
11:43:42.538 SPE "o"
11:43:43.177 LOC =D3 [ ]
11:43:43.178 SPE "r"
11:43:44.137 LOC =J4 [ ]
11:43:44.138 SPE "l"
11:43:44.777 LOC =D4 [ ]
11:43:44.778 SPE "d"
11:43:49.097 LOC =J5 [ ]
11:43:49.257 LOC =J5 [ ]
11:43:49.417 LOC =J5 [ ]
11:43:49.657 LOC =J5 [ ]
11:43:50.057 LOC =J5 [ ]
11:43:50.937 LOC =I3 [ ]
11:43:50.938 SPE "o"
11:43:51.648 LOC =D3 [ ]
11:43:51.650 SPE "r"
11:43:52.457 LOC =J4 [ ]
11:43:52.458 SPE "l"
11:43:53.257 LOC =D4 [ ]
11:43:53.258 SPE "d"
11:43:53.897 LOC =E6 [ ]
11:43:53.906 SPE " "
- Something may be in this backup https://www.assistiveware.com/support/proloquo4text/protect-customizations/export-to-dropbox
From Brian:
"Extracting just strings from an obl file should be pretty easy. it's just JSON and you'd do the following pseudo-code:
session['events'].each |event|
if event['label']
// only button events have labels
end
end
end
obla files would have gibberish in the label field for any words that weren't considered "core" for that user"
Message History can be exported as a plain text file
You cant export message history
??