At the end of a project, you upload project documents to a Microsoft SharePoint Online library that contains many files. The following is a sample of the project document file names:
✑ aei_AA989.docx
✑ bci_WS098.docx
✑ cei_DF112.docx
✑ ebc_QQ454.docx
✑ ecc_BB565.docx
All documents that use this naming format must be labeled as Project Documents:
You need to create an auto-apply retention label policy.
What should you use to identify the files?
To identify files based on a specific naming format in Microsoft SharePoint Online, creating a custom sensitive information type (SIT) is the most appropriate approach. This method involves using regular expressions to detect patterns in filenames, making it ideal for identifying project documents with a consistent naming format. While trainable classifiers are useful for content analysis, they are more complex and unnecessary for simple filename pattern recognition.
I think this should be A. A sensitive info type, a regular expression kan detect filename (a-z + d{3}) easy and fast.
I tested and a sensitive info type is not able to verify the file name.
All methods checks for content. In this case we don't know the content, so the best thing to do is use trainable classifiers to check the content and find a pattern to identify the files. In my opinion, the question includes files names just for confuse us!
SIT doesn’t check file names, only content. It has to be a trainable classifier.
Sensitive info type does not classify based on file name
My first thought was what, that's not true, but indeed it only checks the content not the file name.
Sensitivity info type makes more sense to me, as it's focused on just the project documents in a site with "many files". If you do a trainable classifier you'd need a large set of "good" examples and this could take a while. But if you were to create a custom sensitive info type and use regex expressions then it should be simpler to implement.
SIT verify file names or only content of that files ?
From Chat GPT: To automatically apply retention labels based on specific criteria, such as file names, you would typically use a trainable classifier in Microsoft 365. Trainable classifiers in Microsoft 365, specifically sensitive information types, can be trained to recognize patterns in data, including file names, and apply actions based on the identified patterns. In this scenario, you would create a trainable classifier that recognizes the naming format of the project documents (e.g., files with names like "aei_AA989.docx"). Once trained, you can use the trainable classifier to create an auto-apply retention label policy that automatically applies the label to documents that match the specified naming pattern. So, the correct answer is: C. A trainable classifier
Answer = A what they want is a NAMING format, NOT the inside of the file contents as classifer is looking at contents
Agree, I did a test and the filenames are also detected by a sensitive info type.
Question #4 actually contains the answer ;) Question #4Topic 1 HOTSPOT - You use project codes that have a format of three alphabetical characters that represent the project type, followed by three digits, for example Abc123. You need to create a new sensitive info type for the project codes. How should you configure the regular expression to detect the content?
"How should you configure the regular expression to detect the content?" - Content, not file name.
C is correct
The answer is C - trainable classifier.
Like everyone mentioned, SIT doesn't look at file names. Tested in my tenant.
C. A trainable classifier: Trainable classifiers are used for unstructured content and require machine learning to identify files based on examples of similar content. This approach is more suitable for text-based content rather than predictable naming conventions.
I don't think so. For me A is better answer, even C should also work. Name of file have easy pattern, so sensitivi info type based on RegEx will be best choice for me
Why A? Sensitive info type is for sensitive content.
a senstive info type might do but project is organisation specific thus a trainable classifier is correct anwser
C is the correct answer. Everyone if focusing on the ###_ll### and not the simple .docx. Trainable classifiers work with .docx files. (It's that simple.) https://docs.microsoft.com/en-us/microsoft-365/compliance/classifier-learn-about?view=o365-worldwide
I think he has chosen the trainable classifier as it is mentioned in the question: "you upload project documents to a Microsoft SharePoint Online library that contains many files." So the requested files are existing only in SharePoint Library. So I believe the answer is right
Trainable classifiers work with .docx files. https://learn.microsoft.com/en-us/microsoft-365/compliance/classifier-learn-about?view=o365-worldwide&viewFallbackFrom=o365-worldwid
Since we have a bunch of documnents with content that we are going to look for we can create an auto-apply label policy that will use a trainable classifier that will use documents as seed documents to find them in the future.
the question says that the documents are uploaded to a Microsoft SharePoint Online library One of the requirements for trainable classifiers is: Place the seed content in a SharePoint Online folder that is dedicated to holding the seed content only. Make note of the site, library, and folder URL. https://learn.microsoft.com/en-us/microsoft-365/compliance/classifier-get-started-with?view=o365-worldwide Since the upload can easily be used to train the classifier since many examples are in it. I agree with C.
Because the question says: All documents that use this naming format must be labeled as Project Documents I chose SIT - A
Wow, what a tricky question. Really a hard one. From all I've read, I found out that DLP policy applied to Exchange online has the option to check if file name matches a pattern. https://learn.microsoft.com/en-us/purview/dlp-policy-reference SIT definitely does not identify file names. Hence, I would also opt for trainable classifier.
SITs match content within documents but can't be used to identify by file name patterns That's where trainable classifiers come in as they detect based on keywords, metadata or content Article on how Trainable classifiers detect things: https://learn.microsoft.com/en-us/purview/trainable-classifiers-learn-about Automated pattern-matching These categorization mechanisms include finding content using: Keyword or metadata values (keyword query language)