AN OVERVIEW OF AUTOMATED REASONING

被引:6
|
作者
POST, S
SAGE, AP
机构
[1] PLANNING RES CORP,MCLEAN,VA
[2] GEORGE MASON UNIV,SCH INFORMAT TECHNOL & ENGN,FAIRFAX,VA 22030
来源
关键词
D O I
10.1109/21.47822
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
There are many approaches to the study of imperfect-information processing. There are also many ways in which information can be imperfect. It can be uncertain, incomplete, imprecise, inconsistent, ambiguous, or some combination of these. Although a multiplicity of approaches exists, no single approach is comprehensive and no consensus of the requirements and principles for reasoning with incomplete and uncertain information has arisen. In fact, there is vigorous disagreement over fundamental issues, such as whether approaches should be based on logic, measures, or some combination of the two, and whether the measures, if used, should pertain to probability, confidence, belief, or degree of membership in a set. There is even some disagreement as to whether methods other than classical logic or probability theory are needed at all, and whether they are even correct. First discussed are the two general approaches to reasoning with imperfect information: nonmonotonic reasoning and a calculus of uncertainty. Default reasoning is posed as an approach that is potentially capable of integrating many facets of these two approaches. Practical requirements for default reasoning are then established. This is done by identifying a number of cases that involve incomplete and uncertain information and showing how they can be addressed by default reasoning. Parametric and symbolic reasoning are differentiated and it is shown that both types are necessary. This distinction is important and most approaches tend to neglect either the parametric and symbolic aspect of default reasoning, thereby restricting use of the default reasoning to one of the two approaches just discussed. It is established that the functional requirements for default reasoning by identifying five capabilities that are necessary to effect default reasoning. Finally, the major characteristics for systems that handle incomplete and uncertain, as well as other types of imperfect information, are established. These characteristics are important because they correspond to design choices in a model for default reasoning, and they also serve as a means to compare the various approaches with each other. © 1990 IEEE
引用
收藏
页码:202 / 224
页数:23
相关论文
共 50 条
  • [1] Automated reasoning
    Gavanelli, Marco
    Mancini, Toni
    [J]. INTELLIGENZA ARTIFICIALE, 2013, 7 (02) : 113 - 124
  • [2] AUTOMATED REASONING
    WOS, L
    [J]. AMERICAN MATHEMATICAL MONTHLY, 1985, 92 (02): : 85 - 92
  • [3] Applications of automated reasoning
    Furbach, Ulrich
    Obermaier, Claudia
    [J]. KI 2006: ADVANCES IN ARTIFICIAL INTELLIGENCE, PROCEEDINGS, 2007, 4314 : 174 - +
  • [4] The flowering of automated reasoning
    Wos, L
    [J]. MECHANIZING MATHEMATICAL REASONING: ESSAYS IN HONOUR OF JORG H SIEKMANN ON THE OCCASION OF HIS 60TH BIRTHDAY, 2005, 2605 : 204 - 227
  • [5] AUTOMATED REASONING - THEORY
    HUMPERT, B
    [J]. HELVETICA PHYSICA ACTA, 1986, 59 (6-7): : 1264 - 1264
  • [6] AUTOMATED REASONING - APPLICATIONS
    HUMPERT, B
    [J]. HELVETICA PHYSICA ACTA, 1986, 59 (6-7): : 1264 - 1264
  • [7] Automated Reasoning in the Wild
    Furbach, Ulrich
    Pelzer, Bjoern
    Schon, Claudia
    [J]. AUTOMATED DEDUCTION - CADE-25, 2015, 9195 : 55 - 72
  • [8] METHODS OF AUTOMATED REASONING
    BIBEL, W
    [J]. LECTURE NOTES IN COMPUTER SCIENCE, 1986, 232 : 171 - 217
  • [9] Automated Reasoning for Mathematics
    Avigad, Jeremy
    [J]. AUTOMATED REASONING, IJCAR 2024, PT I, 2024, 14739 : 3 - 20
  • [10] Automated Reasoning and Learning for Automated Payroll Management
    Dumancic, Sebastijan
    Meert, Wannes
    Goethals, Stijn
    Stuyckens, Tim
    Huygen, Jelle
    Denies, Koen
    [J]. THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2021, 35 : 15107 - 15116