SupplementaryText with repeating text is being identified as IrrelevantElement about sec-parser HOT 6 CLOSED

Thank you for the feedback.
I added the five missing supplementary text elements in the expected-structure-and-text.json.

Scores
Previous accuracy scores:

New accuracy scores:

Do I push the changes directly to the main branch of the sec-parser-test-data repository?

I appreciate your feedback.

from sec-parser.

Do I push the changes directly to the main branch of the sec-parser-test-data repository?

yes (or create a PR, both options are ok)

from sec-parser.

Hey @Elijas,

I started working on this issue.

For 10-Q MSFT-0000950170-23-014423:
The actual-structure-and-text-summary Json file reports that there are five unexpected SupplementaryText elements.

 "unexpected_elements_by_type": {
            "TopSectionTitle": 10,
            "TitleElement": 18,
            "SupplementaryText": 5
        }

I looked over the five unexpected SupplementaryText elements and they all have the text content "Refer to accompanying notes."
The "SupplementaryTextClassifier" classifies elements that their normalized text expresses the words "accompanying notes" as SupplementaryText elements.

Should I edit the expected-structure-and-text Json file to include them? Or are these supposed not to even be supplementary text elements?

I am taking a more look into the issue and will let you know when I am ready for feedback.

from sec-parser.

Just add these 5 manually to expected-structure-and-text.json and that will be it 👍

from sec-parser.

Pull request: alphanome-ai/sec-parser-test-data#2

from sec-parser.

Thanks! Merged

from sec-parser.