THE DIATHESIS NEWSPAPER DIGITIZATION SUITE

1
THE DIATHESIS NEWSPAPER DIGITIZATION SUITE

• Foundation of Research and Technology
• Institute of Computer Science
• Centre for Cultural Informatics

Heraklion, Crete, Greece
Martin Doerr, Georgios Markakis, Maria Theodoridou
2
About DIATHESIS

• Diathesis is a newspaper digitization suite whose
  primary purpose is the digitization,
  classification and dissemination of archival
  newspaper material.
• It was originally used for the digitization of
  the Vikelaia Municipal Librarys newspaper
  collection (1890-1960) at Heraclion, Crete. It
  has evolved as an independent digitization suite
  since.
• Used in other projects as well (Filekpedeytiki
  Etairia Athens, Greece The AYGHI newspaper)

3
The Problem

• Historical newspapers are one of the most
  signicant source of information for researchers
  due to the wealth of information they provide
  regarding every aspect of everyday political,
  social and intellectual life.
• Access to this type of archival material is
  usually obstructed by the following factors
• In order to protect the archival material from
  potential damage some archives prohibit the
  access to the largest part of their collection.
• Direct contact with the original archival
  material constitutes a potential health hazard
  (due to dust and fungi).
• The lack of indexes to newspapers combined with
  the vastness of information contained in them
  makes research a very time consuming task.
• Many archives adopted digitization of newspapers
  as a straightforward method to deal with the
  above problems. Digitized material is easier to
  preserve and much easier to distribute via the
  Web.
• However, conversion of archival material into a
  digital image format (i.e. JPEG, TIFF, PDF or
  DJVU) does not solve the problem of rapid access
  to this material.
• Digitization itself is inadequate if it does not
  provide the means of rapidly accessing the
  digitized material in a timely and accurate
  manner (also known as the searchability issue).

4
Current State of the Art newspaper Digitization
Practices

• Currently there are three main approaches for
  rendering newspaper archival material searchable
• The Physical Features Based Approach
• The OCR Based Full Text Indexing Approach.
• The Conceptual Classification (Ontology Based)
  Approach.

5
The physical features based classification
approach.

• Newspapers are classified using a basic set of
  metadata regarding physical features of the
  original material (number of issue, date of
  publication, newspaper name, number of pages
  etc).
• Advantages
• Simple to implement.
• Disadvantages
• The final user is unable to conduct full-text
  searches on an article or issue level basis.
• The final outcome of the digitization effort
  resembles more a browsing mechanism.
• There is no explicitly defined conceptual
  structure of the archive.
• Institutions
• Anno Austrian newspapers online project
  (http//deposit.ddb.de/online/exil/exil.htm).
• Exilpresse digital. deutsche exilzeitschriften
  1933-1945" project (http//deposit.ddb.de/online/e
  xil/exil.htm).
• Denmark Digitaliserede danske aviser 1759-1865
  (http//www.statsbiblioteket.dk).

6
The OCR based Full Text Indexing Approach.

• Automatic digitization approaches that make use
  of OCR analysis of digitized newspapers. Full
  Text Indexing techniques are currently considered
  to be the state of the art in the area of
  newspaper digitization and this is mainly for the
  following reasons
• Creation of searchable full - text index via OCR
  is a much faster process compared to the manual
  creation of metadata.
• Separation of searchability and readability.
• It is possible to conduct searches at a
  page/issue/article level basis.
• The search is conducted via keywords in a manner
  that is familiar to the average user of
  contemporary Web Search engines.
• Efficient content dissemination over the Web.
• Disadvantages
• Well known precision/recall issues.
• Newspaper archives are not as chaotic as the Web.
• The search of information in OCR based
  information retrieval systems is conceptually
  blind.
• The import process a computationally expensive
  procedure.

7
The OCR based Full Text Indexing Approach.

• Institutions adopting this approach
• British library online newspaper archive
  (http//www.uk.olivesoftware.com/).
• The Brooklyn Daily Eagle online
  (http//www.brooklynpubliclibrary.org/eagle/).
• Northern New Nork historical newspapers
  (http//news.nnyln.net/).
• Utah Digital Newspapers (http//www.lib.utah.edu/d
  igital/unews/).
• Historical newspapers in Washington
  (http//www.secstate.wa.gov/history/newspapersname
  .aspx).
• To mention just a few

8
The conceptual classification approach.

• The conceptual classification approach overcomes
  many of the above weaknesses by enabling the user
  to perform a knowledge engineering task upon the
  already digitized material via the use of
  ontologies.
• An Ontology "the specifcation of ones
  conceptualization of a knowledge domain".
• Advantages
• Ontologies are used to express a specific
  conceptual view over the digitized material.
• The use of top level ontologies guarantees to a
  certain extent the semantic interoperability
  among different archives.
• The user may use concepts that classify the
  document that are not initially contained within
  the document itself.
• Disadvantages
• Given the density of information in a newspaper,
  production of metadata is a notoriously time
  consuming task (knowledge engineering
  bottleneck).
• It is almost impossible to manually define all
  the semantic relations or entities contained even
  in a single article in a timely manner.

9
The DIATHESIS Approach a hybrid approach

• This system attempts to implement a realistic
  conceptual classification approach by combining
  the best elements from the three approaches
  mentioned above
• It permits searches on a newspaper issue basis
  (newspaper issue name, number, publication date)
  in a similar manner to the physical features
  based approach.
• It permits searches on an article level basis via
  the use of full text queries in a similar manner
  to the OCR based Full Text Indexing Approach.
• It permits searches on an article level basis via
  the semantic relationships assigned to each
  segment.
• It permits searches that combine all of the above
  elements.
• The system DOES not attempt to create a complete
  semantic structure that includes all the semantic
  relationships and entities (Actors, Places)
  described in the text. Instead it focuses to the
  creation a coherent semantic backbone that can be
  easily enriched with semantic relations.
• DIATHESIS is using CIDOC CRM as an underlying
  ontology.

10
Aims of DIATHESIS

• To render the digitized newspapers searchable on
  a document/article level basis.
• To exploit the use of OCR technology in order to
  enable full text search in a newspaper
  collection.
• To combine full text search with user-defined
  metadata based search on a document and article
  level basis in order to enhance the overall
  precision factor of the system.
• To provide visualization facilities and an
  ergonomic interface for
• The timely completion of metadata according to a
  set of predefined thesauri hierarchies.
• The browsing of the digitized newspaper
  collection given a set of predefined thesauri
  hierarchies.
• To deal with issues of semantic interoperability
  of digitized material (conformance to
  international standards).
• To create a robust semantic backbone that will
  allow the full implementation of the CIDOC CRM
  Model.

11
About CIDOC

• What is the CIDOC Conceptual Reference Model?
• An Object Oriented Ontology of about 80 classes
  and 130 properties for cultural and natural
  history
• CRM instances can be encoded in many forms
  RDBMS, ooDBMS, XML, RDF(S), OWL.
• Accepted as ISO-21127 in June 2005
• The CRM
• Is not a metadata standard
• It is meant to become our language for semantic
  interoperability,
• It is a Conceptual Reference Model for analyzing
  and designing cultural information systems
• Is limited to the underlying semantics of
  database schemata and document structures used in
  cultural heritage and museum documentation
• Does not define the terminology used to document
  these data structures
• Does not say what cultural institutions should
  document
• Aims to explain the logic of what they actually
  do document

12
(No Transcript)
13
An Example Hierarchy E70 Stuff (Thing)
14
CIDOC Example (1) Modeling an Activity
February 1945
P82 at some time within
P7 took place at
P11 participated in
E7 Activity
Crimea Conference
P86 falls within
P67 is referred to by
E65 Creation Event

P81 ongoing throughout
P14 performed
P94 has created
15
CIDOC Example (2) Describing a composite
artifact
16
CIDOC-CRM DIATHESIS implementation
Issue/Segments Relationships
17
CIDOC-CRM DIATHESIS implementation Issue
Physical Features
18
CIDOC-CRM DIATHESIS implementation Activity
References
19
Thesauri Hierarchies
20
CIDOC based newspaper annotation
CIDOC CRM Core Ontology
Integration by Factual Relations
Donald Johanson
Discovery of Lucy
Johanson's Expedition
real world nodes (KOS)
Lucy
Hadar
Ethiopia
Benaki Museum
Documents in Digital Libraries
21
The System Architecture Software Components
Apache Tomcat Application Server
Newspaper Digitization Suite
Diathesis Administrator
Diathesis Annotation Mechanism
DIATHESIS Web Search
Database
SIS-TMS Thesaurus Management System
Client Side
Server Side
22
The System Architecture Workflow View
23
The user interface

• FEATURES
• Fully Web Based.
• Simple to use / Easy to learn.
• Intelligent Upload / Download Mechanism.
• Workflow Control .
• Data Loss Prevention Mechanism (Temporary Local
  Storage and Data Recovery).
• Flexible and Ergonomic Completion of Metadata
  Fields.
• Automatic Highlighting of keywords in OCR Text
  (Actors, Places).
• Use of SVG thesauri hierarchies for the timely
  completion of Vocabulary Reserved Metadata fields.

24
The user interface
DIATHESIS
End User Search Mechanism
Annotation Mechanism
Administrator
Search for Subjects
Usage Stats
Search for Issues
Mass Import
System Configuration
25
Demonstration Annotation Interface
26
Demonstration End User Search Mechanism
27
Future Directions

• Enrich the metadata creation process with
  Information Extraction Techniques.
• Expand the suite with complementary Deep Semantic
  Annotation Capabilities (Semantic Wiki)

PHASE 1
PHASE 2
PHASE 3
DIATHESIS Semantic Wiki
Information Extraction Techniques
Material Preprocessing Phase
Shallow Semantic Annotation metadata production
phase.
Deep Semantic Annotation full CIDOC
implementation phase
28
Conclusions

• The use of OCR technology in newspaper
  digitization practices is a hot new technology.
  However it is not capable to deal with a plethora
  of issues.
• Deep Semantic annotation via Semantic Web
  technologies is a promising future trend. CIDOC
  CRM provides the theoretical means to achieve
  this. The problem is how to implement it.
  Creation of deep semantic relationships that
  exist within the boundaries of a single newspaper
  issue is a time consuming , and therefore
  expensive task.
• The DIATHESIS digitization suite encapsulates a
  digitization strategy towards the creation of a
  vast semantic network of factual relationships
  between CIDOC entities while effectively dealing
  with the following issues
• Digitization and Storage of Newspaper Material
• Rendering digitized material searchable on an
  issue/article level basis via the use of
  metadata, thesauri hierarchies and full text
  queries.
• Create a semantic backbone that can be used by
  future implementations.
• The next step Link the DIATHESIS semantic
  backbone with a Semantic Wiki.

29

• Thank You!

geomark_at_ics.forth.gr martin_at_ics.forth.gr