Abstract Abductive logic programming is a logical representation of abductive reasoning. Most ALP frameworks express domain-specific logical relationships that the abductive answers must satisfy. A priori, the integrity constraints are known. However, for some applications (e.g., Interactive Abduction Logic Programming, Multi-Agent Interactions, Contracting), it is reasonable to loosen this assumption so that the abductive reasoning process starts incompletely aware of the... Abductive logic programming is a logical representation of abductive reasoning. Most ALP frameworks express domain-specific logical relationships that the abductive answers must satisfy. A priori, the integrity constraints are known. However, for some applications (e.g., Interactive Abduction Logic Programming, Multi-Agent Interactions, Contracting), it is reasonable to loosen this assumption so that the abductive reasoning process starts incompletely aware of the integrity constraints and continues without restarting when a new integrity constraint is known. In the present paper, we provide an abstract syntax for the declarative logic programming of abductive logic, with the addition of integrity constraint during the process of abductive reasoning, an operational implementation with formal termination, good and completeness properties, and an implementation based on SCIFF language and proof procedure. Read More Read Less

Abstract It is introducing a new non-monotonic ILP ( NILP ) approach, and its implementation is called top-direct abductive learning (TAL). TAL addresses some shortcomings of inverse-entailed ILP systems and is the first Top-Down ILP system to allow background theories and hypotheses to be normal logic programs. Tal’s approach is based on mapping an ILP problem to an equivalent ALP one.... It is introducing a new non-monotonic ILP ( NILP ) approach, and its implementation is called top-direct abductive learning (TAL). TAL addresses some shortcomings of inverse-entailed ILP systems and is the first Top-Down ILP system to allow background theories and hypotheses to be normal logic programs. Tal’s approach is based on mapping an ILP problem to an equivalent ALP one. This allows the use of well-established ALP proof procedures and the specification of richer language biases with integrity constraints. The mapping provides a principled search space for the ILP problem and allows an abductive search to compute inductive solutions. Read More Read Less

Abstract Higher-order property comes into play in some aspects of climatological impact research. For instance, vulnerability measures, which are essential in determining the vulnerability to climate change of different regions and places, must satisfy certain conditions best expressed by quantifying the overall increasing functions of the appropriate type. This type of property is often considered to be “cognitive”, but for the... Higher-order property comes into play in some aspects of climatological impact research. For instance, vulnerability measures, which are essential in determining the vulnerability to climate change of different regions and places, must satisfy certain conditions best expressed by quantifying the overall increasing functions of the appropriate type. This type of property is often considered to be “cognitive”, but for the measures used in practice, it is relatively easy to code the property as a dependent type and prove it correct. In scientific programming, it is common to care about the “correctness” of the program up to the “implicit”: for example, the program would perform as expected, for example, if real numbers were used instead of floating point values. These “counterfactuals” are impossible to test, but they are easy to code and prove as types. We show examples (encoded in AGDA) encountered in actual vulnerability assessment. Read More Read Less