Transcript Conclusions - Kniha ve 21. století

The possibility of using
methods and techniques
of artificial intelligence
in the library activities
Jolanta Szulc
Institute of Library and Information Science
University of Silesia in Katowice, Poland
KNIHA VE 21. STOLETÍ, Opava, 2016
Objective
• The aim of the presentation is to discuss the possibility
of using methods and techniques of artificial intelligence
in library activities.
• To achieve this objective the analysis of literature
accumulated on the basis of queries in selected
databases: LISA and LIST was conducted.
Content
• Searching and selecting scientific literature
• Definitions and criteria for the classification of artificial
intelligence
• Examples of applications of artificial intelligence in
libraries
• Conclusions
Searching and selecting
scientific literature
LISA: Library and Information
Science Abstracts
Search term
Number of records
Artificial intelligence AND Library work
779
Date of publication
Number of records
1970-1979
19
1980-1989
207
1990-1999
204
2000-2009
262
2010-2016
86
Source type
Number of records
Scientific journals
779
LISA (access: 15.02.2016)
Publication title
Number of records
Artificial Intelligence
95
AI Magazine
61
Artificial Intelligence in Medicine
49
Artificial Intelligence Review
49
Artificial Intelligence and Law
30
IEEE Intelligent Systems
27
AI Communications
17
……
…..
Library Hi Tech
15
……
…..
Kniznice a Vedecke Informacie
……
Small Computers in Libraries
……
5
…..
2
…..
LISA (access: 15.02.2016)
Publication type
Journal Article
Number of records
778
Bibliography
1
Book Review
1
Conference Report
1
LISA (access: 15.02.2016)
Subject
Number of
records
artificial intelligence
414
information work
247
information storage and retrieval
193
subject indexing
183
technical services
165
computerized information storage and retrieval
163
searching
128
……
…..
……
…..
cataloguing
11
librarianship
11
……
…..
libraries
5
online cataloguing
5
LISA (access: 15.02.2016)
Classification
Number of
records
COMPUTER APPLICATION
51
COMPUTER SCIENCE
23
AUTOMATIC TEXT ANALYSIS, AUTOMATIC INDEXING, MACHINE
TRANSLATION
19
MANAGEMENT OF COMPUTING AND INFORMATION SYSTEMS (CI)
8
INFORMATION WORK
7
COMPUTERIZED INFORMATION STORAGE AND RETRIEVAL
6
……
CLASSIFICATION
……
LIBRARIANSHIP AND INFORMATION SCIENCE
…..
…..
4
…..
2
…..
LIBRARY TECHNOLOGY - NETWORKS
1
RESEARCH
1
LISA (access: 15.02.2016)
Language
Number of records
English
696
Russian
23
German
11
Japanese
11
French
10
Czech
5
Swedish
4
Chinese
3
Italian
2
Polish
2
Portuguese
2
Slovak
2
Spanish
2
……
…..
LISA: Library and Information
Science Abstracts
(access: 15.02.2016)
Search term
Number of records
artificial intelligence library
4281
Date of
publication
Number of records
Source type
Number of records
1960-1969
1
Journal article
1970-1979
30
Book review
19
1980-1989
300
Conference report
9
1990-1999
1742
Feature
2
2000-2009
1716
Bibliography
1
2010-2016
481
4273
LISTA: Library, Information
Science & Technology Abstracts
(access: 15.02.2016)
Search term
Number of records
Artificial intelligence AND Library work
13
Date of publication
Number of records
1970-1979
0
1980-1989
4
1990-1999
4
2000-2009
2
2010-2016
3
Source type
Number of records
Magazines
8
Academic journals
4
Books
1
Definitions and criteria
for the classification of
artificial intelligence
Definitions and terms of AI (1)
1) Sztuczna inteligencja to nauka polecania maszynom
wykonania zadań wymagających inteligencji, jaką użyłby
człowiek przy wykonaniu identycznych zadań. (Minsky
1968)
2) Sztuczna inteligencja jest tą częścią informatyki, która
bada procesy rozumowania symbolicznego i
niealgorytmicznego oraz zajmuje się reprezentacją
symbolicznie ujętej wiedzy. (Forsyth 1984)
3) Sztuczna Inteligencja to ekscytujące próby stworzenia
myślących komputerów […], maszyn z umysłami w
pełnym tego słowa znaczeniu. (Haugeland 1985)
4) Sztuczna Inteligencja jest reaktywowanym podejściem
do rozwiązywania problemów, które wymaga
wewnętrznego, heurystycznego, poszukiwania w czasie
operacji na bazie wiedzy, zwłaszcza dla logicznych i
analogicznych mechanizmów rozumowania. (Glover,
Greenberg 1989)
Definitions and terms of AI (2)
5) Sztuczna Inteligencja to dziedzina badań, podczas
których próbuje się naśladować ludzką inteligencję w
maszynie. Obszar AI obejmują systemy oparte na
wiedzy, systemy eksperckie, rozpoznawania obrazów,
automatyczną naukę, rozumienie języka naturalnego,
robotykę i inne. (Kurzweil 1999)
6) Sztuczna Inteligencja jest to nauka o czynnościach, które
miałyby spowodować, że maszyny będą wykonywać
funkcje, które aktualnie lepiej wykonuje człowiek. (Rich,
Knight 1990)
7) Sztuczna Inteligencja jest to dziedzina nauki próbująca
wyjaśnić i emulować inteligentne zachowania za pomocą
metod obliczeniowych. (Schalkoff 1990)
Definitions and terms of AI (3)
8) Sztuczna inteligencja [podejmuje] prace nad metodami
obliczeniowymi, które umożliwiałyby [maszynom]
postrzeganie, wnioskowanie, działanie. (Winston 1992)
9) Termin Artficial Intelligence, w skrócie AI, odnosi się do
komputerów, które naśladują aspekty ludzkiego
myślenia. Prosty kalkulator elektroniczny nie ma statusu
AI. Ale maszyna, która może uczyć się na swych
błędach, albo to może popisać się rozumowaniem, ma
status AI. Między tymi skrajnościami, nie ma żadnej
dokładnej linii podziału. (McGraw 1994)
10) Dziedzina Sztucznej Inteligencji (AI, Artificial
Intelligence) – obejmuje rozwiązywanie problemów
sposobami wzorowanymi na naturalnych działaniach i
procesach poznawczych człowieka, za pomocą
symulujących ją programów komputerowych
(Kwaśnicka 1990)
Criteria for the classification of AI (1)
Due to the lack of universally accepted criteria for the
distribution of artificial intelligence suggested its division
into subareas due to two criteria:
• solving class of tasks,
• technique used.
Criteria for the classification of AI (2)
The following areas of AI are distinguished (according to
the first criterion):
• automated theorem proving,
• automatic programming,
• computer vision,
• expert systems,
• game theory,
• information retrieval (intelligent database)
• machine learning ,
• natural language processing (NLP),
• problem solving and strategies searches,
• processes of perception,
• robotics .
Criteria for the classification of AI (3)
The following areas of AI are distinguished (according to
the second criterion):
• Case-Based Reasoning (CBR)
• Expert Systems (ES),
• Fuzzy System (FS),
• Genetic Agorithms,
• Intelligent Information Systems (IIS),
• Neural Network (NN).
Examples of applications
of artificial intelligence
in libraries
Semantic Geo-Catalogue (SGC) project
The enhancement of the search capabilities of geo-spatial
tools occupies one of the highest positions in the agenda
of the INSPIRE (INfrastructure for SPatial InfoRmation
in Europe) initiative.
The paper presents work on the Semantic Geo-Catalogue
(SGC) project in providing a semantic extension to the
geo-catalogue of the Autonomous Province of Trento
(PAT) in Italy.
•
F. Farazi, V. Maltese, B. Dutta, A. Ivanyukovich, V. Rizzi: A semantic
geo-catalogue for a local administration. „Artificial Intelligence Review”
Vol. 40 (Aug 2013), p. 193-212
Semantic Geo-Catalogue (SGC) project
Expert systems in library and
information science
Methods: An analysis was made of the literature, gathered
on the basis of queries in some databases.
The article discuses directions of research in the field of
expert systems as one of the methodologies of
knowledge management, using methods of artificial
intelligence: problem solving and representation of
knowledge (generation, valuation and knowledge
management, extraction and synthesis of knowledge in
the construction of expertise), "self learning", evaluation
of performance. In conclusion, it attempts to identify
possible areas of applications in libraries and information
centers.
•
J. Szulc: Systemy ekspertowe w działalności bibliotecznej i
informacyjnej: stan badań, problemy badawcze, przykłady zastosowań.
„Zagadnienia Informacji Naukowej. Studia Informacyjne” Vol. 52, no. 1
(103) (2014), s. 94-118.
Expert systems in library and
information science
Expert systems in library and
information science
Expert systems in library and
information science
Expert systems in library and
information science
Examples of expert systems:
• ESSCAPE (Exeter of University , UK, 1982-1983, )
• MAKLUM (Malaysian, 1996)
• DOCMATCH (University of Bradford, UK, 1996)
Expert systems in library and
information science
Conclusions:
Quantitative analysis of the literature showed a significant
decrease in the literature devoted to expert systems
after 2010. Nevertheless expert systems (experimental,
operated) are used in many fields. It is expected that
these systems will be used in specialized libraries in the
collection and codification of knowledge in selected
areas of specialization.
Other applications of AI:
• Application of intelligent agents in libraries
• Chatbots in the library
• Semantic research for digital libraries
Conclusions
Conclusions
• Analysis of the literature showed the maintenance of a
relatively large number of publications on AI
• Relatively few article relates to uses of AI in libraries
• Scientific literature was grouped in categories such as:
automatic text analysis, automatic indexing, computer
application, computer science, machine translation and
other
• Examples of applications of AI relate to uses of such
fields of library and information science as: classification,
library technology, research, subject indexing and other
• Interdisciplinary character of the research is to be
noticed – the disciplines penetrate and complement each
other
Conclusions
• Applicability AI in different types of libraries is still not
fully successful. Various methods and techniques AI can
greatly assist the process of ordering and information
flow.