Several authors have pointed out some limitations inherent in Montague's approach to natural language representation, and have attempted to remedy these limitations by working out suitable extensions of the Montague formalism. One of these limitations originates from the fact that logical connectives and modal operators can only be applied to formulae and not to arbitrary logical expressions, which stands in contradiction with the common practice of natural language where connectives and modalities can affect expressions of various syntactic categories (and not only complete sentences). This paper introduces an extension of Montague's intensional logic that precisely addresses the problem just alluded to. The proposed extension is equipped with a well-defined 'Boolean semantics', analogous with the Keenan-Faltz semantics.

Covers some of the most significant applications of artificial intelligence, namely: natural language processing, speech understanding, expert system design, requirement engineering, machine learning, truth maintenance systems, advanced concepts and methods of logic programming. Together with the previous two volumes edited by Thayse, this completes a comprehensive exposition of the subject of logics applied to AI.

Naturallanguage processing constitutes the subfield of artificial intelligence which deals with the development of computer programs able to understand natural language. Montague's semantics and Boolean semantics attempt to apply some techniques of mathematical logic and of algebraic lattice theory respectively for the representation of natural language with a view to its automatic processing. The purpose of th is paper is to compare these two approaches from syntactic and semantic points of view.

Knowledge representation natural and formal languages modal logic intensional logic and Montague's semantics temporal logic and specification of concurrent programs revisable reasoning theory of relational and deductive databases representation of incomplete information in databases.

Theorem in proving and P-functions.- Grammars, logics and declarative programming.- Grammars and Semantics.

This paper shows that synthesizing binary decision programs (formed by means of decision instructions of the type if then else and of execution instructions of the type do) and proving theorems can be carried out by using the same approach. It is proved that the same transformations acting on P-functions can be interpreted in terms of binary program synthesis and of theorem proving. Since binary program leads to algorithmic state machine design while theorem proving leads to declarative programming, this allows us to lay a bridge between logic design and declarative languages such as Prolog.

We show that well-known instructions such as if then else, fork, join, while do, can be represented as row matrices or column-matrices. We define a matrix-instruction which encompasses and generalizes the above instructions. This instruction provides us with a compact tool for describing algorithms and for synthesizing them in synchronous and asynchronous structures. We show, e.g., that the synthesis of a program by means of elementary instructions reduces to the factorization of a matrix into elementary matrices. A formalism and a computation method are introduced which generalize the author's previous work on the subject.

A formalism has been introduced for program description and synthesis, namely the matrix description of instructions. In this paper we put that formalism to work by associating with it computation methods based on a generalized P-function concept. Algorithms are derived for the optimal implementation of programs in asynchronously organized structures.

The main objective of recommender systems is to help users select their desired items, where a major challenge is modeling users’ preferences based on their historical feedback (e.g., clicks, purchases or check-ins). Recently, several recommendation models have utilized the adversarial technique, which has been successfully used to capture real data distributions in various domains (e.g., computer vision). Nevertheless, the training process of the original adversarial technique is very slow and unstable in the domain of recommender systems. First, the sparsity of the implicit feedback dataset aggravates the inherently intractable adversarial training process. Second, since the original adversarial model is designed for differentiable values (e.g., images), the discrete items also increase the training difficulty. To cope with these issues, we propose a novel method named Adversarial Pairwise Learning (APL), which unifies generative and discriminative models via adversarial learning. Specifically, based on the weaker assumption that the user prefers observed items over generated items, APL exploits pairwise ranking to accelerate the convergence and enhance the stability of adversarial learning. Additionally, a differentiable procedure is adopted to replace the discrete item sampling to optimize APL via backpropagation and stabilize the training process. Extensive experiments under multiple recommendation scenarios demonstrate APL’s effectiveness, fast convergence and stability. Our implementation of APL is available at: https://github.com/ZhongchuanSun/APL.

A program is defined as an indexed sequence of instructions; each of these instructions is formed by an interconnection of branching (or conditional) instructions (of the form if, then, else) followed by an interconnection of execution instructions (of the form do). A program is an efficient tool, allowing the digital system designer to describe the microprograms of discrete systems and to synthesize their control automaton. This paper deals with a method of transformation and of optimization of programs. The presented algorithm obtains, for any given program, an equivalent one with a minimum number of conditional vertices.

This paper considers the realization of switching functions by programs composed of certain conditional transfers (binary programs). Methods exist for optimizing binary trees, i.e., binary programs without reconvergent instructions. This paper studies methods for optimizing binary simple programs (programs with possible reconvergent instructions, but where a variable may be tested only once during a computation) and binary programs. The hardware implementations of these programs involve either multiplexers or demultiplexers and OR-gates.

This paper considers the realization of algorithms by programs composed of conditional instructions (binary programs). A mathematical tool, the algebra of P-functions, is introduced for the analysis and synthesis of these programs. P-functions present the following advantages with respect to other known methods: they detect automatically the reconvergent instructions and are able to synthesize a class of programs (non-simple programs) non reachable by other methods. 1. Introduetion The purpose of the present paper is to introduce the concept of P-function which will appear to be the adequate mathematical tool for synthesizing, transforming and hence optimizing evaluation programs. Section 2 may be considered as a large introductory chapter which simultaneously states the problems at hand and describes in a formal way the types of solutions that will be proposed. In section 3 one develops the algebra of P-functions which will be used in section 4 for synthesizing programs and in part IU of this work for optimizing these evaluation programs. (1) 2. Introduetion to conditional program transformation 2.1. Problem statement Glushkov 1,2) has proposed to describe systems performing computations, a model of cooperation between two subsystems, called operational automaton and control automaton, respectively (see ref. 3). The control automaton is the actual implementation of the computation algorithm; it accepts instructions of the following type. N

A new method for obtaining the test functions for detecting all the stuck faults that can occur at the inputs of a combinational switching circuit is presented. This method is based on the use of the Taylor expansions of the Boolean function realized by the switching circuit.