Previous Topic: Features No Longer SupportedNext Topic: FSA Rescans Unmodified Items After Upgrade


Text Extraction By Autonomy KeyView

CA DLP now uses Autonomy KeyView technology to extract the text content of captured events. (Previous versions of CA DLP used Oracle Outside In technology.)

With the transfer to KeyView technology, there may be some differences in the extracted data compared to previous CA DLP releases. However, our testing indicates that any effects on the accuracy of existing user policies are negligible. Indeed, in some cases, the transfer to KeyView technology is likely to result in fewer false positives.

Impact on Fingerprinted Documents

(Only affects customers upgrading from CA DLP 12.5)

Content agent triggers created in CA DLP 12.5 continue to work when you upgrade to CA DLP 14.0. In general, any existing content agent triggers continue to detect documents that were fingerprinted using 12.5 content agents.

However, existing content agent triggers may fail to detect PDFs and, especially, spreadsheets that were fingerprinted using 12.5 content agents. This problem mainly affects Text Detection content agents with sentence- or paragraph-level accuracy. The problem is caused by minor discrepancies between Outside In and KeyView text extraction technologies. Specifically, the upgraded content agent triggers now use KeyView technology to analyze documents. But these triggers are comparing the documents against a content index of document signatures that was generated using Outside In technology.

To ensure that fingerprinted spreadsheets and PDFs are reliably detected after you upgrade to CA DLP 14.0, we recommend that you regenerate the affected content agents. Rebuild the content indexes and then republish the content agents.