- The biggest limitation of our tool is the amount of data that is stored in our databases and the amount of self-training and user feedback that has been given to our tool.
- An example of this is the following sentence, "this is good, but it ask too many questions." If the tool has seen "it asks" in the crawled data and it also recognized it as a "verb" it can identify that it should be "asks" not "ask". Also, the same thing about "questions". If the tool hasn't seen "questions" a noun it might suggest a typo correction to "question".
- The tool is unable to detect the language and it is assumed that the user is giving the desired language by specifying the target.
- The corrector is mostly effective for typo correction. There are some hyper paramters within the code that can be tuned to help the corrector decide how aggresive it should rewrite the code, but still it has known problem with suggesting a replacement.
- Both corrector and checker are limited by how large the crawled data was and how much the user fix them.
- The corrector is unable to fix the relation of groups of word more than 2 because it's not using the n-grams.
- The way Apostrophe s is handled is to expand the clause, e.g. "it's -> it is" and "John's car -> car of John"
- If for any reason the database left in-use by a process and you kept getting the error that an specific `databse is locked`, run `fuser databaseName.db` and kill all the process that are active, using `kill -9 <process_number>`
- The crawled data was not enough so we also used some prompt given texts using LLMs.
#### Android App
- If a dialog box pops up and the device/screen is rotated, then the app will unexpectedly crash
