DIGIBOOK

1
DIGIBOOK

• Mervin John
• Lindsey Cameron
• Cynthia Balloch
• Dan Curran
• Rich Przekop
• ES96 Spring 2004
• Prof. Abernathy
• Prof. Yang

2
Digibook Review

• Objective To Create a User Interface for the
  Gordon McKay Book Collection that builds on
  HOLLIS
• Primary Features
• Scanned Images of Cover, TOCs, Index
• Virtual Browsing

3
The Virtual Bookshelf

• Objective With most, if not all, books slated
  for the depository, there must exist some
  on-site, off-site virtual browsing capability
• HOLLIS provides significant browsing and search
  capabilities but does not attempt to replace
  shelf browsing
• I.e browsing by section and call number (w/
  title, and book info)

4
Book Hierarchy

• Sort books into expandable hierarchy based on the
  Library of Congress Classification Hierarchy

Library
McKay
Class
T - Technology
Sub-Class
TK Electrical Engineering
TK7885-TK7895 Computing Hardware
Sub-Division
CMOS Digital Design
Book
5
My Virtual Bookshelf

• Similar to online Commercial Shopping Carts
• Allows users to add/remove books that they are
  currently browsing
• Gives current status of book (location,
  checked-out, on-site/off-site)

6
Shelf Browsing

• Search Results will give books along with scanned
  image of covers w/ links to TOCs and Indexes
• Browse surrounding environment
• /- 5, /-10, /-25
• Look at shelves above and below
• Look at entire subject

7
Current Concerns

• How we can integrate with Hollis system?
• How will it integrate with the website?
• What facilities will be shared between different
  projects?
• Programming language semantics
• Goal is to have these issues worked out within
  the next week

8
HOLLIS Information

• Contact Paul Aloisio aloisio_at_gentoo.harvard.edu
• Can search database with barcodes using CGI link
• http//holliscatalog.harvard.edu/F/?funcfind-cCC
  L_TERMbarTHE_BAR_CODE
• Better searches can also be done using HOLLIS
  numbers (can restrict to specific library MCK)
• http//holliscatalog.harvard.edu/F/?funcitem-glob
  aldoc_libraryHVD01doc_number002510780yearvo
  lumesub_libraryMCKtype02
• Several other possibilities Endnote, Z39.50
  search tools?

9
The Process

• Photocopy title-page, TOC, and index
• Scan title-page/TOC and index as two separate
  files
• ReadIRIS OCR software produces searchable PDF
  text file
• 0-2 errors per page, words still readable

10
Figures

• 6 min/book to copy and scan (on average)
• Roughly 10 books/hour
• Student employee working at 9/hour
• Approximately 0.90/book

11
Other Factors

• Adjustable frame improves efficiency, makes
  copies more uniform
• Additional time needed to get book off shelf,
  transit between library-copier-scanner
• Paid by the hour, not by the book

12
Automation

• Goal 1 click creation of each books webpage
• Online form to upload TOC and index, and barcode
  reader
• http//dlib.deas.harvard.edu/digibook/save_book.ht
  ml
• Problem with SFTP and uploading

13
Searching

• Use MNOGO to browse books
• Able to search text of pdf files
• http//dlib.deas.harvard.edu/samples/
• Can control the results page to fit our needs
• Can re-index at any time

14
Storing Information