Page 1

40 Years Gesellschaft f¨ ur Klassifikation – From Numerical Taxonomy to Big Data – Hans-Hermann Bock 1. 2. 3. 4. 5.

The origins Activities - Conferences - Topics Publication activities Persons and working groups Aspects for the future

GfKl Jubilee Colloquium, Karlsruhe, February 24, 2017 GfKl40-2017 - 0.1


GfKl was founded on February 12, 1977 at Frankfurt Main Station 8 founding members: Dr. Ingetraut Dahlberg Dr. Robert Fugmann Dr. Hans-Hermann Bock Dipl.-Volksw. Otto Gekeler Prof. Dr. Martin Scheele Dr. Hans-Georg Sch¨on .....

Philosophy, documentation (DGD) Chemical databases (Hoechst) Mathematics, statistics (TU Hannover) Commodity classification (BWL, AEG) Biological taxonomy/documentation Librarian (UB M¨unster)

GfKl40-2017 - 1.1


Classification in 1960-1977 • Broad interest in classification methods

• Societies:

Classification and indexing for libraries (Dewey Decim. Classif., Thesauri) Classification systems in documentation e.g., chemical documentation Universal classification (of all entities ...) Product and commodity classificaton Patent classification Index Medicus Biological taxonomy and systematics Mathematical methods, classification rules, clustering algorithms Analyse des donn´ees

Gesellschaft f¨ur Dokumentation (DGD) Gesellschaft f¨ur Information und Dokumentation (GID) F´ed´eration Intern. de Documentation (FID) Committee ’Classification Research’ International Society for Commodity Classification and Technology (IGWT) Infoterm (Vienna) Systematics Association (USA, UK) Classification Society (USA, UK)

GfKl40-2017 - 1.2


R.R. Sokal, P.H.A. Sneath Principles of Numerical Taxonomy 1965

I. Dahlberg Grundlagen universaler Wissensordnung 1974

H.-H. Bock Automatische Klassifikation 1974

GfKl40-2017 - 1.3


GfKl40-2017 - 1.4


2. Activities - conferences - topics

First conference in M¨ unster, June 4, 1977 with 44 participants and 9 papers: – Linguistic bases of scientific conceptualization (Gipper) – Conceptual bases of classification (Dahlberg) – A dichotomous classification of ’definitions’ (Nacke) – Usage of the term ’concept’ (Nacke) – Reproduction of concept relations by a classification – Fundamentals of numerical classification (Bock) – Bases of ordination and seriation (Ihm) – Classification principles in libraries (Lohse)

GfKl40-2017 - 2.1


400

325 324

300

312

330

250

220

206

230 250

260 259 216 210 192 215

150

102

114 102

93

100

114

64 70

81

28

21

31

28

41

70 54

74

83

92

115 130 95

143

146

132

153

128

125

122

146

165

180

202 200

189

200

144 27 43

9

8

Berlin Magdeburg Dortmund Cottbus Mannheim München 2001 Passau 2000 Bielefeld 1999 Dresden 1998 Potsdam 1997 Freiburg 1996 Basel 1995 Oldenburg 1994 Kaiserslautern 1993 Dortmund 1992 Salzburg 1991 Marburg 1990 zugleich 1.IFCS-Conference

Augsburg 1989 Darmstadt 1988 Aachen 1987 Münster 1986 Karlsruhe 1985 Hofgeismar 1984 Königswinter 1983 Augsburg 1982 Hofgeismar 1981 Salzburg 1980 Königstein/Ts.1979 Frankfurt/Höchst 1978 Münster 1977

1977 12.2.

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

0 0

126

100

365

300 300 293 291

126 112

100 70

50 50

350

318 308 302

315

312

326

331 318

277

300 266 245

200 165 136

150 134

150

205 209 190

200

250 233

250

350 302

300

Vorträge Teilnehmer 350

The conferences: # participants/lectures Number of members of GfKl

Jahrestagungen der Gesellschaft für Klassifikation

GfKl-07 - 2.2


From the ’conceptual period’ to ’statistical data analysis’: 1977 - 1989 First topics in GfKl: Order and concept theory Philosophical bases Universal classification Linguistic methods Documentation Thesaurus construction Automatic indexing Library classification Numerical classification

from about 1984 =⇒

The new topics: Exploratory data analysis Discriminant analysis, pattern recognition Clustering models and algorithms Multidimensional scaling, visualization Correspondence analysis Formal concept analysis (lattice theory) Preference structures Many applications of numerical methods

⇓ 1989: New Society splits off: International Society for Knowledge Organisation (ISKO)

GfKl40-2017 - 2.3


International relationships 1 • 9th GfKl conference in Karlsruhe (1985, W. Gaul)

with many internationally renowned experts: Arabie, Carroll, Critchley, De Leeuw, Hubert, Murtagh, De Soete, Wishart GfKl40-2017 - 2.4


July 1985, Cambridge, UK: Founding the International Federation of ClassiďŹ cation Societies (IFCS)

July 1987, Aachen, Germany: 1st Conference of the IFCS

GfKl40-2017 - 2.5


Following the methodological trends at conferences: GfKl Conference 1998 (Dresden): Statistical data analysis ClassiďŹ cation and clustering Formal concept analysis Learning algorithms Internet and information technology Usage of new media and internet Information systems ClassiďŹ cation for multimedia, WWW Navigation and classif. on the internet Virtual universities

Applications Marketing, market research Finance, capital, risk management Bioinformatics, Genomics,... Environment Health sciences Archaeology

GfKl40-2017 - 2.6


GfKl Conference 2007 (Freiburg) Statistical data analysis – Multidimensional data analysis – Symbolic data analysis – Data visualization – Time series and panel data – Image and signal analysis Classification – Supervised classif., pattern recognition – Cluster analysis – Similarity structures – Mixture analysis, latent class models – Classification and regression Data mining – Analyzing data streams – Text mining, web mining – Semantic web – Statistical relational learning – Knowledge representation and discovery

Marketing and management science – Banking and finance – Data analysis in retailing – Business intelligence and personalization Various applications – Biostatistics and bioinformatics – Medical and health sciences – Statistical musicology – Linguistics – Archaeology and archaeometry – Library classification, automatic indexing

GfKl40-2017 - 2.7


International relationships 2 Since 2013: • GfKl conferences are organized as ”European Conference on Data Analysis” (ECDA) •

related to the foundation of a ”European Association for Data Science” (EuADS)

incorporated in Luxembourg

ECDA

Place

2013: 2014:

University of Luxembourg Jacobs-Universit¨at Bremen

Cooperating societies

French Classification Society SFC Italian CLADAG, Dutch VOC, Polish SKAD, IASC Univ. of Essex, Colchester SKAD and British Classification Society BCS 2015: GfKl-2016: DAGStat, G¨ottingen GfKl, Biometric Society/DR, ... 2016: European Data Science Conf. EuADS in Luxembourg-Grund 2017: Wroclaw, Poland Polish SKAD, European EuADS, Italian CLADAG, British BCS

GfKl40-2017 - 2.8


39th Annual Conference of GfKl: European Conference on Data Analysis - ECDA-2015 Colchester/UK Data analysis and statistics (8) – Supervised classification (2) – Clustering (3) – Optimization methods (1) – Data analytics (1) – Crowdsourcing (1) – Machine learning and knowledge discovery (4) – Big data (1) – Math. foundations of Data Science (2)

Applications – Economics and finance (3) – Marketing (2) – Engineering, logistics (1) – Biostatistics, bioinformatics, life sciences (1) – Digital humanities, social sciences (2) – Education (1) – Archaeology and geosciences (1) – Musicology (1) Library classification, automatic indexing

GfKl40-2017 - 2.9


International relationships 3:

IFCS

Network of IFCS Societies: SKAD CLAD CSNA

GfKl

IPRCS

BCS JCS VOC SFC CLADAG

KCS Slov

• Organization of international IFCS conferences: 1987 Aachen, 2009 Dresden, 2011 Frankfurt • Biennual bilateral ’Symposia on Data Analysis and its Applications’ – GPSDA with Polish Classification Society (SKAD) Aachen 2009, Krakau 2011, Dresden 2013 – GJSDA with Japanese Classification Society (SKAD) Tokyo 2005, Berlin 2006, Karlsruhe 2010, Kyoto 2012, Reisensburg 2016 • Special joint conferences, e.g.: – with Italian CLADAG: Firenze 2010 GfKl40-2017 - 2.10


3. Publications activities

Studien zur Klassifikation 1 (1977) - 19 (1989)

International Classification 1 (1974) - 19 (1992) GfKl40-2017 - 3.1


Studies in ClassiďŹ cation, Data Analysis and Knowledge Organization Since 1989, thus far: 50 volumes

Advances in Data Analysis and ClassiďŹ cation (ADAC) 1 (2007) - 10 (2017)

GfKl40-2017 - 3.2


Since 2013/14: Archives of Data Science - Open Access publications of GfKl Editorial Board: A. Geyer-Schulz, E. H¨ullermeier, H. Kestler KIT - Scientific Publishing www.ArchivesforDataScience.org

GfKl40-2017 - 3.3


4. Persons and working groups The presidents of GfKl

1998-2004

Claus Weihs 2004-2013

Berthold Lausen since 2013 GfKl40-2017 - 4.1


Working Groups: Datenanalyse und Numerische Klassifikation (AG-DANK)

G. Ritter

Chr. Hennig

Since 1978

Hans-J. Mucha

Datenanalyse und Klassifikation in den Biowissenschaften (AG-BT) Since 1987: Biologische Taxonomie; P.O. Degens, W. Ludwig, B. Lausen, H. Kestler

P.-O. Degens

P.O. Degens

B. Lausen

H. Kestler GfKl40-2017 - 4.2


Computeranwendungen und Quantitative Methoden in der Arch¨ aologie (AG-ARCH, CAA) since 1990, chaired by P. Ihm †, A. Zimmermann, I. Herzog, K. Lambers

Bibliotheken, Bibliotheksklassifikation (AG-BIB) since 1979: H. D¨uvel,...., F. Scholze

Dezimalklassifikationen (AG-DK) since 1992, chaired by G. Knell, A. Weiße, H.-J. Hermes, B. Lorenz

GfKl40-2017 - 4.3


Specific topics: Formale Begriffsanalyse

(Rudolf Wille † )

Linguistik, Dialektometrie (Hans Goebl)

Economics

(Gaul, Opitz, Decker, Baier,...)

Musicology (Claus Weihs) GfKl40-2017 - 4.4


The current Board of GfKl

GfKl40-2017 - 4.5


5. Aspects for the future: Data Science and Big Data Major tasks for GfKl – Attracting young scientists – New problem fields, methods and applications – Data Science – External profiling Big Data – Role of experimental design – Differing types of validity of results – Ethical guidelines

GfKl40-2017 - 5.1

40 Jahre GfKl: Vortrag von Prof. Dr. Dr. Bock  
40 Jahre GfKl: Vortrag von Prof. Dr. Dr. Bock  
Advertisement