FOA Home
The AI Geneology project (AIG) is an attempt to collect information
relating authors in artificial intelligence to one another through
shared advisors. In analogy to geneological family trees, we can treat
the advisor as parent to advisee. Students of students become
grandchildren, etc. A subset of this data is shown in Figure
(FOAref) . As an example, of how this additional information
about authors can be exploited to understand more about the words used
in documents, Steier has analyzed word-pair co-occurrance data as a
means of finding potential index phrases [REF1097] . Take, for example, the phrase
{\tt CASE-BASED REASONING}. This phrase has a high degree of phrasal
information content (using mutual information statistics to identify
statistically-dependent word pairs) when these statistics are collcted
across the entire AIT document corpus.
\small McCulloch, W. S. ----
Minsky, Marvin ---- ---- Winston, Patrick H. ---- 1973 Waltz, David L.
---- 1980 Finin, Timothy W. THE SEMANTIC INTERPRETATION... 1988 Kass,
Robert John ACQUIRING A MODEL OF THE... 1989 Klein, David A. SEE CO-ADV
SHORTLIFFE, TED 1987 Pollack, Jordan B. ON CONNECTIONST MODELS OF...
1993 Angeline, Peter EVOLUTIONARY ALGORITHMS AND... 1994 Kolen, John
COMPUTATION IN RECURRENT...
Mey, Jacob 1969 Schank, Roger C. A CONCEPTUAL
DEPENDENCY... ---- Kolodner, Janet L. ---- 1989 Shinn, Hong Shik A
UNIFIED APPROACH TO... 1989 Turner, Roy Marvin A SCHEMA-BASED MODEL
OF... 1991 Hinrichs, Thomas Ryland PROBLEM-SOLVING IN OPEN... 1992
Redmond, Michael Albert LEARNING BY OBSERVING AND... ---- Dejong, Gerald
Francis ---- 1987 Segre, Alberto Maria EXPLANATION-BASED LEARNING...
1988 Shavlik, Jude William GENERALIZING THE STRUCTURE... 1991 Towell,
Geoffrey Gilmer SYMBOLIC KNOWLEDGE AND... 1988 Mooney, Raymond Joseph A
GENERAL EXPLANATION-BASED... 1992 Ng, Hwee Tou A GENERAL ABDUCTIVE
SYSTEM... 1989 Rajamoney, Shankar Anandsubramaniam EXPLANATION-BASED
THEORY... ---- Lehnert, Wendy G. ---- 1983 Dyer, Michael G. IN-DEPTH
UNDERSTANDING: A... 1987 Mueller, Eric DAYDREAMING AND COMPUTATION:...
1987 Zernik, Uri STRATEGIES IN LANGUAGE... 1988 Pazzani, Michael John
LEARNING CAUSAL... 1988 Gasser, Michael SEE CO-ADV HATCH, EVELYN 1989
Dolan, Charles Patrick TENSOR MANIPULATION... 1989 Alvarado, Sergio Jose
UNDERSTANDING EDITORIAL... 1989 Dolan, Charles THE USE AND ACQUISITION
OF... 1991 Lee, Geunbae DISTRIBUTED SEMANTIC... 1991 Reeves, John
Fairbanks COMPUTATIONAL MORALITY: A... 1991 Nenov, Valeriy Iliev
PERCEPTUALLY GROUNDED... 1991 Quilici, Alexander Eric THE CORRECTION
MACHINE: A... 1993 Turner, Scott R. MINSTREL: A COMPUTER MODEL... 1990
Williams, Robert Stuart LEARNING PLAN SCHEMAS FROM... 1976 Meehan, James
R. THE METANOVEL: WRITING... 1978 Wilensky, Robert UNDERSTANDING
GOAL-BASED... 1985 Jacobs, Paul A KNOWLEDGE-BASED APPROACH... 1986
Norvig, Peter A UNIFIED THEORY OF... 1986 Arens, Yigal CLUSTER: AN
APPORACH TO... 1987 Chin, David Ngi INTELLIGENT AGENTS AS A... 1992 Wu,
Dekai AUTOMATIC INFERENCE: A... 1978 Carbonell, Jaime G. SUBJECTIVE
UNDERSTANDING:... 1988 Minton, Steven LEARNING EFFECTIVE SEARCH... 1989
Lehman, Jill Fain ADAPTIVE PARSING:... 1991 Perlin, Mark W. AUTOMATING
THE CONSTRUCTION... 1991 Hauptmann, Alexander Georg MEANING FROM
STRUCTURE IN... 1992 Veloso, Manuela M. LEARNING BY ANALOGICAL... 1980
Lebowitz, Michael GENERALIZATION AND MEMORY IN... 1987 Hovy, Eduard
Hendrik GENERATING NATURAL LANUGAGE... 1988 Hunter, Lawrence E. GAINING
EXPERTISE THROUGH... 1989 Ram, Ashwin QUESTION-DRIVEN... 1989 Dehn,
Natalie Jane COMPUTER STORY-WRITING: THE... 1990 Leake, David Browder
EVALUATING EXPLANATIONS 1992 Domeshek, Eric Andrew DO THE RIGHT THING:
A... 1993 Edelson, Daniel Choy LEARNING FROM STORIES:... \caption{A
sample of the AIGenealogical record}
But a statistically-significant
different distribution is observed within dissertations coming from a
particular set of universities: Yale, Georgia Tech. and the University
of Massachusetts. Within these particular university contexts, the
constituent concepts of { CASE-BASED} and {\tt REASONING} are examined
in detail and independently. Not only do these words occur together as
{\tt CASE-BASED REASONING}, but they often occur separately (e.g. {\tt
CASE-BASED PLANNING, CASE-BASED ARGUMENT, CASE-BASED PROBLEM,
SCHEMA-BASED REASONING, ANALOGICAL REASONING, COMMONSENSE REASONING}).
Within the limited context of these sub-corpora, the keywords occur more
independently and hence are considered less phrase-like. As this phrase
is ``exported'' into the general AI vocabulary, the semantic nuances are
left behind, and the dominant use of the constituent words is as part of
the phrase.
But what is it that makes language use so much different at
Yale, Georgia Tech. and the University of Massachusetts? Perhaps it is
merely coincidence, but our hypothesis is that it is the common lineage
traced to Roger Schank! Work across these geographically distant
research institutions is pulled together by an intellectual tradition
captured by the AIG data.
Part of what is interesting about the AIT
corpus is its demonstration within a single corpus of many of the
representations discussed in this chapter. The AI genealogy captures
some of the intellectual linkage relations arising from the Ph.D.
advisor/advisee relationship. David Waltz' taxonomy of AI provides an
excellent initial thesaurus over keywords [REF331] .
Top of Page
A.I. Geneology