QUESTIONS_AND_ANSWERS.docx

QUESTIONS AND ANSWERS FOR DOCUMENT IDENTIFICATION PROCESS (DIP)

SOLICITATION 28321318R00000022

1. Is there a page limit for any given volume, or for the three volumes combined?

Answer: No, there is no page limit. The 5 MB ceiling is for each electronic submission (email), if your proposal is over 5 MB, you may send more than one.

Answer: Yes.

4. For Volume 2, Section 508 Compliance, may offerors embed VPATs as PDF files within a Microsoft Word document for easier readability?

Answer: Yes.

5. Are the images to be compared black and white only?... or color?

Answer Images are both black and white and color. Majority is black and white.

6. Do the comparisons include PDF-to-TIFF (could the same page exist in the library as both TIFF and PDF?) ... or will the task be limited to comparing PDF-to-PDF and TIFF-to-TIFF, when looking for duplicates?

Answer: Comparison will be PDF to PDF and Tiff to Tiff.

7. Can we assume that all images to be compared are of the same dimensions (e.g. 8.5"x11" page images?).

Answer: Majority is 8.5” x 11”. However, we accept source documents of varying sizes.

8. It states that "The DIP system shall support processing 100K documents per day (real-time), averaging 50 pages per document; however single documents can exceed 2000 pages...". The quantities should be no problem. However, we need to understand the documents: Q: In the case of these multi-page documents, is this a single multi-page document that needs to be compared to all other documents of the same size?... or is it the case that a single "50-page document" may actually be 50 individual 1-page documents, each of which needs to be compared to all other 1-page documents, to check for duplicates?

Answer: In this use case example it is unlikely that documents being compared would be 50 individual 1 page documents. The typical use case would be a document containing multiple pages being compared against another document with multiple page.

9. The attachment A pricing tables show a lot of hours -- yet the overall goal seems to be COTS software that can identify and remove the duplicate images automatically. Q: Please clarify if these are "Managed Services" hours to use the software, or what these hours will be used to cover?

Answer: The manage service hours will be used to support installing, configuring and integration work efforts. The hours used in integration activities would be used to support integrating the COTS software into SSAs infrastructure and document datastore.

10. Identical images can be spotted automatically. However, slight differences may appear in page images (e.g., due to differences in how they were scanned, skew/speckle, etc.) Such very similar but not quite identical pages could be shown to a human, who could be the final arbiter of whether two pages were, in fact, duplicates. Q: Is the SSA looking for a fully-automated process that finds pages are not "identical", but almost certainly different images of the same page? Or, by contrast, would the SSA prefer a workflow process that showed very similar pages to a human, who could then be the final arbiter and confirm the pages as duplicate, or non-dup pages?

Answer: In this use case we are looking for a fully-automated process to find duplicate documents. We are also looking for the capability for the end user to be the final arbiter and confirm pages are a duplicate. A viewer would enable the user view the pages and take an action to identify the page as a duplicate.

11. Are there certain regions of the page, which are mission-critical in determining duplicates? (Sometimes a particular region of a page contains mission critical information, such as SSN or applicant address, which be inspected carefully looking for duplicates, without having to inspect the entire page in detail... and with millions of pages to check against, this can speed up comparisons a lot).

Answer: We receive many types of documents that do have common regions that potentially could be used for detection and pattern analysis (example SSA forms).

12. How many connectors will be required for this project? ie. SharePoint, etc.

Answer: Assuming connectors are SSA interfaces, the COTS product would need to support APIs that could be used to connect and interoperate with SSA’s document architecture, electronic folder and security infrastructure.

13. How many/what applications does SSA require integration with?

Answer: The COTS product would need to support APIs that could be used to connect and interoperate with SSA’s document architecture, electronic folder and security infrastructure.

14. How many images will SSA expect to pass through the software application being proposed/used per year?

Answer: The DIP system shall support processing 100K documents per day on average (averaging 50 pages per document). The system will need the ability to scale to 700,000 documents, averaging 50 pages per document as additional growth is expected.

15. Will SSA create/accept an additional CLIN for Consulting? If not, can we expect to combine into Integration/developer hours?

Answer: No we do not want a separate category. You must try and incorporate into the labor hours and categories provided.

16. Without a specific scope, a definitive amount of hours cannot be determined. Please clarify that SSA will accept a baseline installation and an hourly rate on the proposal for additional hours?

Answer: We will not accept a baseline installation, however, after careful review, the base year pricing for services will be changed from fixed price to labor hours with the labor hours being provided in the pricing tables in Section A.