Executive Summary of Digital LibraryDigital Museum at NCTU

1
Executive Summary of Digital Library/Digital
Museum at NCTU

• 1Hao-Ren Ke and 2Wei-Pang Yang
• Library / Department of Computer and Information
  Science,
• National Chiao-Tung University, Hsinchu, Taiwan,
  R.O.C.
• 1claven_at_lib.nctu.edu.tw, 2wpyang_at_cis.nctu.edu.tw

2
Introduction

• DL_at_NCTU Architecture
• An infrastructure for
• Representing digital objects, capacities of DLs
• Supporting the interoperability of distributed
  DLs
• Metadata technology -- MML
• VUCS_at_NCTU -- Case study of MML
• PIE_at_NCTU -- Personalization Information
  Environment
• Customization, effective search, tracking user
  interests
• YuYu Yang Digital Museum
• A digital museum for the masterpieces of an
  internationally well-known sculptor, YuYu Yang

3
DL_at_NCTU Architecture
4
DL_at_NCTU

• Research on digital libraries at NCTU
• Four goals of DL_at_NCTU
• Build an infrastructure for the interoperability
  of digital libraries
• Identify appropriate DL services according to
  users needs
• Support the exchange of information in diverse
  formats
• Study the economic aspect in DL
• The DL_at_NCTU architecture
• 3-Tier architecture Metadata Manager, Registry
  Authority, and Content Repository
• Kernel metadata technology

5
Metadata in Digital Libraries

• Metadata
• Structured data about data, such as Dublin Core,
  GILS
• Machine understandable information about
  resources or other things
• An aid to resource discovery in the digital
  library community
• Research issues for metadata in digital libraries
• Data representation and description
• Content management

6
3-Tier Architecture of DL_at_NCTU

• Metadata Manager takes responsibilities for
  metadata manipulation, service allocation and
  content query
• Schema Editor, Metadata Editor, Translation
  Editor
• Registry Authority manages the information of
  digital objects, autonomous DL services, and
  metadata
• Registry Table, Service Table, Content Table
• Content Repository is the repository of digital
  objects, schema, and metadata
• Handle is generated by Content Repository

7
Conceptual Model of DL_at_NCTU Architecture
Presentation/User Interface
Metadata Editor
Metadata Manager
Translation Editor
Schema Editor
Data Representation Content Management
Content Repository
Registry Authority
8
Conceptual Model for DL_at_NCTU Metadata
1-3 Metadata Manager 4 RA and CR
9
Features of DL_at_NCTU Metadata

• Container-based metadata
• Define a language for modeling digital objects
  capacities of autonomous DL services -- MML
• Provide a translation mechanism to bridge
  interoperability gap -- also by MML
• A flexible architecture for content management

Metadata is the kernel of DL_at_NCTU
10
Metadata Modeling Language (MML)

• Rich modeling power for describing data formats,
  service capabilities, and translation rules
• A translation mechanism between metadata to
  achieve the interoperability of autonomous DLs
• Encapsulate data in object format and container
  technology
• Adapt XML as language syntax and the simplified
  data model revised from RDF with further
  extending its functions
• Two parts of MML
• Data Model
• Translation Service

11
MML Data Model

• MML data model adapts the ideas of E-R model and
  Object-Oriented Model
• Three import parts of MML Data Model Entity,
  Attribute and Value
• The basic unit of MML is Resource
• A resource can be associated with other resources
  as an Attribute
• An attribute has a Value
• Both Attribute and Value are also resources

Attribute (Resource)
Value (Resource)
Entity (Resource)
12
MML Data Model

• Basic Data Model
• Hierarchical Data Model

13
MML Schema
Target Database
Target Schema
MML Schema
Target Metadata
MML Metadata
Translation Rule Metadata
Metadata Translation (for interoperability)
Translation Rule Schema
Schema Class in OO
MML Schema
Source Metadata
Source Database
Source Schema
MML Schema
14
MML Metadata
Target Database
Target Schema
Target Metadata
MML Metadata
Translation Rule Metadata
Metadata Translation (for interoperability)
Translation Rule Schema
MML Metadata
MML Metadata
Source Metadata
Metadata Instance in OO
Source Database
Source Schema
15
Interoperability of Heterogeneous Metadata
16
MML Data Model
Attribute (Resource)
Value (Resource)
Entity (Resource)
MML Metadata
Resource A unit with ten basic properties (ISO
11179 - Specification and Standardization of Data
Elements)
17
MML Syntax
XML Syntax

• MML ltmmlPrefixgt
  resources lt/mmlPrefixgt
• mmlPrefix Model Model_Instance
• resources properties
  attributeItems stringvalue
• properties ltpropertyTaggt
  propertyValue lt/propertyTaggt
• propertyTag MML_Name MML_Type
  MML_Doc Identifier
• MML_Version
  MML_RA MML_Language
• MML_Definition
  MML_Obligation MML_MO
• propertyVaule string integer float
  boolean stringvalue
• attributeItems ltattributeTaggt resource
  lt/attributeTaggt
• attributeTag Attribute attributename

18
Example MML Schema (book)

• ltModelgt
• ltMML_NamegtBooklt/MML_Namegt
• ltMML_TypegtBooklt/MML_Typegt
• ltMML_DocgtThis is a model of a booklt/MML_Docgt
• ltIdentifiergtBook_ID1lt/Identifiergt
• ltMML_VersiongtVersion_1lt/MML_Versiongt
• ltMML_RAgtNCTU_DBlt/MML_RAgt
• ltMML_LanguagegtEnglishlt/MML_Languagegt
• ltMML_DefinitiongtUnlimitedlt/MML_Definitiongt
• ltMML_ObligationgtUnlimitedlt/MML_Obligationgt
• ltMML_MOgtUnlimitedlt/MML_MOgt

• ltAttributegt
• ltMML_NamegtTitlelt/MML_Namegt
• ltMML_TypegtStringlt/MML_Typegt
• ltMML_DocgtThe title of a booklt/MML_Docgt
• lt/Attributegt
• ltAttributegt
• ltMML_NamegtAuthorlt/MML_Namegt
• ltMML_Typegt_at_Authorlt/MML_Typegt
• ltMML_DocgtDescribe the writer of a booklt/MML_Docgt
• lt/Attributegt
• lt/Modelgt

19
Example MML Schema (Author)

• ltModelgt
• ltMML_NamegtAuthorlt/MML_Namegt
• ltMML_TypegtAuthorlt/MML_Typegt
• ltMML_DocgtThis is a model of an authorlt/MML_Docgt
• ltIdentifiergtAuthor_IDlt/Identifiergt
• ltMML_VersiongtVersion_1lt/MML_Versiongt
• ltMML_RAgtNCTU_DBlt/MML_RAgt
• ltMML_LanguagegtEnglishlt/MML_Languagegt
• ltMML_DefinitiongtUnlimitedlt/MML_Definitiongt
• ltMML_ObligationgtUnlimitedlt/MML_Obligationgt
• ltMML_MOgtUnlimitedlt/MML_MOgt

• ltAttributegt
• ltMML_NamegtFirst_Namelt/MML_Namegt
• ltMML_TypegtStringlt/MML_Typegt
• ltMML_DocgtThe first name of an authorlt/MML_Docgt
• lt/Attributegt
• ltMML_NamegtLast_Namelt/MML_Namegt
• ltMML_TypegtStringlt/MML_Typegt
• ltMML_DocgtThe last name of an authorlt/MML_Docgt
• lt/Attributegt
• lt/Modelgt

20
Example MML Metadata

• Book Metadata with id Book_ID1
• ltModel_Instancegt
• ltMML_TypegtBooklt/MML_Typegt
• ltIdentifiergtBook_ID2lt/Identifiergt
• ltBookgt
• ltTitlegtPLlt/Titlegt
• ltAuthorgtAuthor_ID2lt/Authorgt
• lt/Bookgt
• lt/Model_Instancegt

• Author metadata with id Author_ID1
• ltModel_Instancegt
• ltMML_TypegtAuthorlt/MML_Typegt
• ltIdentifiergtAuthor_ID2lt/Identifiergt
• ltAuthorgt
• ltFirst_NamegtRavilt/First_Namegt
• ltLast_NamegtSethilt/Last_Namegt
• lt/Authorgt
• lt/Model_Instancegt

21
MML Metadata/Schema Editor
22
MML Translation Service

• MML translation service
• Translate the source metadata into a target
  format, according to the translation rules

Translation Rule Template Template Name
Name_of_Translation_Rule From
From_Attribute Model To
To_Attribute Model Translation Rule
Target_Attribute1 Source Attribute
Translation Rule Operation Target_Attribute2
Source Attribute Translation Rule Operation
23
MML Translation Service

• In the MML translation service, we support
  several basic operations to manipulate MML
  metadata

Basic operation in MML
Operation
Type
Add, Sub, Multiply, divide
Integer
String concatenate
String
Add, Sub, Multiply, divide
Float
And, Or
Boolean
24
ExampleMML Schema (Translation Rule Template)

• ltModelgt
• ltMML_NamegtTranslation_Rule_Templatelt/MML_Namegt
• ltMML_TypegtTranslation_Rule_Templatelt/MML_Typegt
• ltMML_DocgtThis is a template of translation
  servicelt/MML_Docgt
• ltIdentifiergtTranslation_Template_ID1lt/Identifiergt
• ltMML_VersiongtVersion_1lt/MML_Versiongt
• ltMML_RAgtNCTU_DBlt/MML_RAgt
• ltMML_LanguagegtEnglishlt/MML_Languagegt
• ltMML_DefinitiongtUnlimitedlt/MML_Definitiongt
• ltMML_ObligationgtUnlimitedlt/MML_Obligationgt
• ltMML_MOgtUnlimitedlt/MML_MOgt

• ltAttributegt
• ltMML_NamegtFromAttriModelNamelt/MML_Namegt
• ltMML_TypegtStringlt/MML_Typegt
• lt/Attributegt
• ltAttributegt
• ltMML_NamegtToAttriModelNamelt/MML_Namegt
• ltMML_TypegtStringlt/MML_Typegt
• lt/Attributegt
• ltAttributegt
• ltMML_NamegtTranslationRulelt/MML_Namegt
• ltMML_TypegtStringlt/MML_Typegt
• ltMML_MOgtUnlimitedlt/MML_MOgt
• lt/Attributegt
• lt/Modelgt

Translation rules are also defined by MML
25
MML Translation Service
26
Comparison of MML, RDF and Stanford Metadata
27
Case Study -VUCS_at_NCTU

• Virtual Union Catalog System (VUCS)
• Alternative to centralized union catalog system
• Search the WebPAC (Web-based Open Public Access
  Catalog) of participating libraries on the fly
• Techniques used
• Information Extraction from Structured Documents
  with Common Structure
• Documents with identical embedded structure and
  semantics
• Web search engine, WebPAC
• MML
• Modeling the metadata of search results of a
  WebPAC
• Define canonical metadata and translation
  mechanism

28
Architecture of VUCS_at_NCTU
VUCS Service
1. Integrated User Interface
1. Distributed Object Transport 2. MML Canonical
Metadata 3. MML Translation
Distributed Object Transport (CORBA)
1. Extract information from HTML documents 2.
Translate data into MML Native Metadata
VUCS Server
VUCS Server
DL_at_NCTU Metadata
Extract
Homepage
WebPac
WebPac
29
Level-IDs Model

• Level-IDs Model
• Define structure hierarchy for structured
  documents
• Give each important tag a Level-ID

1 ltHTMLgt 1.1
ltTABLEgt 1.1.1 ltTRgt 1.1.1.1
ltTDgtTitlelt/TDgt 1.1.1.2 ltTDgtAuthorlt/TDgt 1.1.2
ltTRgtltTDgt...lt/TDgt 1.1.3
ltTRgtltTDgt...lt/TDgt 1.1.4 ltTRgtltTDgt...lt/TDgt
lt/TABLEgt
lt/HTMLgt
Level-IDs Assign
Level-IDs Model
Structured Documents
Structure Hierarchy
Documents with Level-IDs
Auxiliary Table or DTD
30
Information Extraction For Structured Documents
with Common Structure
31
First Phase - Preprocessing

• Reformulate structured documents into a
  formalized and labeled format
• Formalization
• Tidy A utility created by World Wide Web
  Consortium
• Automatically tidy up sloppy tag into a nicely
  layout markup
• Convenient and accurate for Level-IDs assignment
• Level-IDs assignment

32
Second Phase - Structure Labeling

• Mark iteration items
• Give semantics for each marked item
• Determine the iteration root
• Wrap data into Label File in MML format

33
Structure Labeling Tool Information Extractor
34
Third Phase - Information Extraction

• 1 lttable border"border" cellspacing"0"gt
• 1.1 lttrgt
• 1.1.1 lttdgtltinput typeSubmit
  value?????????????? / ltinput typeRESET
  nameCLEAR value??????" /gtlt/tdgt
• 1.1.2 lttdgt??6?????lt/tdgt
• 1.1.3 lttdgt??/???1/1lt/tdgt
• lt/trgt
• lt/tablegt
• 2 lttable border"1" cellspacing"0"gt
• 2.1 lttr bgcolor"AAAAFF"gt
• 2.1.1 ltthgtlt/thgt
• 2.1.2 ltthgt???(??HyperLink?????)lt/thgt
• 2.1.3 ltthgt??/???lt/thgt
• 2.1.4 ltthgt?????lt/thgt
• 2.1.5 ltthgt???lt/thgt
• 2.1.6 ltthgt????lt/thgtlt/trgt
• 2.2 lttr bgcolor"DEB887"gt
• 2.2.1 lttdgtlta href"../STORE.DLL/MARC?ID2728
  46"gt1ltinput type"checkbox name"L"
  value"272846" /gtlt/agtlt/tdgt
• 2.2.2 lttdgtlta href"../store.dll/?ID272846a
  mpT2ampSASC"gtUnderstanding digital PCS
  the TDMA standard /lt/agtlt/tdgt
• 2.2.3 lttdgtCoursey, Cameron. /Artech
  House,lt/tdgt

Iteration Root 2 Iteration Level 2.1 Iteration
Item 2.1.2, 2.1.3, 2.1.4, 2.1.5, 2.1.6
35
VUCS_at_NCTU
36
VUCS_at_NCTU
37
Personalized Information Environment PIE_at_NCTU
38
??

• ??
• ????
• PIE_at_NCTU
• ?????
• ???????
• ????
• ????????
• ????
• ?????
• ???????
• ????

39
PIE_at_NCTU

• A personalized environment for library services
  and information
• Features
• Personalized and customized user interface
• Dynamically track user interests to help users
  find information
• Selective Dissemination of Information (SDI)
• User Interface
• Personal desktop
• Personal search engine
• Personal bookmark
• Selective dissemination of information (SDI)
• Personal announcement

40
Personal Desktop

• Users can customize the information shown on
  desktop
• Customizable information categorylibrary
  service, WebPAC, databases, E-journals, new book,

41
Personal Search Engine

• Whats the interests of a user?
• Profile
• Query history the query terms used
• Dynamically track user interests
• Frequency of each query term used before
• Inter-relationship among query terms
• Time-discounted (half-decay) weighting for query
  terms and inter-relationship among query terms
• Personal Search Engine
• Rearrange the present query strategy according to
  the user interests
• Reorder the search results according to the user
  interests

42
Keyword Relation Graph -- KRG

• KRG Keep the inter-relationship among query
  terms used before
• Each node is a query term
• If the user uses terms A and B together, a edge
  between node A and B exists
• Associate a weight with each node and edge

Poem
Prose
Classic
Digital Library
43
?????????????

• ???????????
• Wij ???????? i AND j
  ??????????
• ??k??????,???????? i AND j
  ? ?????????
• ?k???????(???)

44
Example
45
Personal Search Engine

• ??????????????
• PIE_at_NCTU????????,????????????????????
• ? (??0.8) ??? (??0.5)
• PIE_at_NCTU????????????????????
• (1) ?and?and??
• (2) ?and?
• (3) ?and??
• (4) ?
• ????????
• ???????????????????????
• ??????????????????
• ?????????????????????,????????????
• ?????????

46
Personal Search Engine

• Query Term poem
• Weighting of poem/prose (0.8), poem/classic (0.5)
• Reorder the search result
• poem prose classic
• poem prose
• poem classic
• poem

47
Personal Bookmark and Announcement, SDI

• Personal Bookmark
• Interesting bookmarks and annotation
• Search strategies
• Personal Announcement
• Related information such as borrowing records,
  overdue records
• Activities interesting to the user
• SDI
• Information filtering
• Send new information to the user according to his
  interests

48
???????

• ?????,??????????,????????????????????(PIE_at_NCTU)
• PIE_at_NCTU???????,?????????,????????????????????
• PIE_at_NCTU??????????????????????????
• ????
• ???????????
• ????
• ????????

49
Taiwan Folk Artist Digital Museum YuYu Yang
Digital Museum
50
Introduction

• Folk Artist
• Folk arts significantly reflect the culture and
  heritage of a country
• Taiwan Folk Artist Digital Museum
• A digital museum with the works of arts created
  by Taiwan folk artists as contents
• Arouse the attention of the public on folk arts
• Originate from an initiative to acquire the
  manuscripts of Taiwan folk artists conducted by
  NCTU
• Its a long-term jobwe choose the works of YuYu
  Yang as the pilot project
• YuYu Yang Digital Museum
• 2000 Theme Project of NSC Digital Museum Program

51
YuYu Yang

• Works -- 1000
• Sculpture stone, marble, stainless, steel,
  bronze
• Painting watercolor, oils, ink
• Printmaking woodcut, lithography, laser printing
• Exhibitions, Collections, and International
  Awards.

52
YuYu Yang (Cont.)
53
Working Group
Dublin Core XML
54
Metadata, Browsing, Searching

• Metadata Dublin Core
• Cataloging Metalogy
• XML DTD
• Browsing/Searching
• Export cataloging results in XML format
• Import into local RDBMS
• Develop browsing and searching functions by ASP

Metalogy
RDBMS
DigitalizedWork
Web Pages Generation(Browsing)
Search Function
55
Esthetic Design and Interactive Functions

• Esthetic Design
• Reflect the style of Yang a natural integration
  of tradition and innovation, of the East and the
  West
• Embed the works of Yang into the Web pages
• Interactive Functions
• Interactive Education
• The creation of a lifescape
• Greeting Cards, Take a Guess
• Screen Saver, News, QA, Bulletin Board

56
Virtual Reality

• Provide users an immersive view of Yangs works
• 3D Model-Based VR
• VRML (Active Worlds)
• Image-Based VR
• 360o Panoramic pictures
• Object-surrounded and Environment-surrounds
• Virtual digital art
• Synthesize Yangs works with virtual backgrounds

57
Web Page Organization

• Introduction
• Preface, Working Group
• About YuYu Yang
• Chronicle
• Turning points of Yang
• Related commentary
• Related paper
• Works of Yang
• Browse by materials
• Search

• Virtual Reality
• 3D Model-based VR
• 2D Image-based VR
• Virtual digital art
• Publication
• Promotion and education
• Interaction and entertainment
• News, bulletin board

58
Folk Artist DM Portal Site
Taiwan Folk Artist DM Portal Site
Trace the development of Taiwan Folk Arts
YuYu Yang
Exhibition
Sen-Yong Su
Integrate -- NCTUs Gallery/ Concert/ Art
Lecture/
Artist Depiction
Artists Philosophy, Style, Masterpiece
Elizabeth Kao
Integrated Search
Interaction
Games, News, Bulletin Board, Opinion
More
Profile/SDI/ Art Passport
Member
59
Personalization

• Member
• User profile art interest, query strategy
• Personalized Service
• Personalized desktop
• Personalized search engine
• Personalized bookmark, annotation
• Auto SDI (Selective Dissemination of Information)
• Art passport
• Interest Group
• Sharable bookmark, annotation, query strategy

60
Conclusions and Future Works
61
Conclusions

• DL_at_NCTU Architecture
• An infrastructure for the interoperability of DLs
  by metadata
• Metadata Modeling Language (MML) to define the
  metadata of digital objects, DL capabilities,
  translation rules
• VUCS_at_NCTU -- Case study of MML
• PIE_at_NCTU
• Customizable user interface, tracking user
  interests for helping personal search
• YuYu Yang Digital Museum
• A digital museum for the masterpieces of an
  internationally well-known sculptor, YuYu Yang

62
Future Works

• DL_at_NCTU Architecture
• Study issues about metadata manipulation service
  query, content query, and service allocation
• PIE_at_NCTU
• Expanded search by using KRG
• Sharing user interests
• Knowledge discovery of user communities
• Taiwan Folk Artist Digital Museum
• More collections
• Portal site

63
(No Transcript)