Reasoning tractably about explicit belief: A model-theoretic approach

The development of the IShopping Agent's knowledge base, the intelligent shopping agent which is named in our research project and proposed the IShopping Agent's knowledge base explicitly takes into account the fact that customers do not adopt the same decision style and that the marketing strategy mix applied by e-commerce companies to promote and sale their product has an influence on the search activity in the sense that their marketing strategy influences the way Web sites are presented. Therefore, using this knowledge base, the Ishopping Agent identifies user's information needs and the most relevant information for customers.

This paper develops two agent knowledge bases in conceptual graph form, one using the KD45 underlying logical model for belief and one without any underlying logical model for belief. Action-attitudes in the knowledge bases provide contexts that represent the agents’ mental attitude towards, and willingness to act upon information in the knowledge bases. Preconditions for communication acts are also represented in the knowledge bases as well as mental attitude changes following communications. Conceptual graphs are a flexible and extendable form of knowledge representation that is used to capture and represent semantic constituents of communications in a form that may be used by software agents. The knowledge base representations in this paper provide software agents a perspective from which they may reason about the communicating agent’s beliefs and communication actions.

We introduce a multi-agent logic of explicit, implicit belief and awareness with a semantics using belief bases. The novelty of our approach is that an agent’s awareness is not a primitive but is directly computed from the agent’s belief base. We prove soundness and completeness of the logic relative to the belief base semantics. Furthermore, we provide a polynomial embedding of the logic of propositional awareness into it.

This paper addresses the dual problems of recognizing when an agent is expressing an attitude of doubt toward a proposition and of realizing natural language utterances that effectively convey such doubt. It presents naturally occurring examples that illustrate the wide variety of ways in which doubt can be expressed. The paper presents a recognition algorithm that uses linguistic, contextual, and world knowledge to determine when an agent is expressing doubt at a proposition and to identify the proposition that is being doubted. It then presents a realization algorithm that uses these knowledge sources to formulate utterances that will convey doubt when an agent cannot accept a proposition due to conflict between it and some other proposition in the agent's knowledge base.

This paper discusses an argumentation system that treats argumentation dynamically. We previously proposed a model for dynamic treatment of argumentation in which all lines of argumentation are executed in succession, with the change of the agent's knowledge base. This system was designed for grasping the behaviour of actual argumentation, but it has several limitations. In this paper, we propose an extended system in which these points are revised so that the model can more precisely simulate actual argumentation. In addition, we provide a simpler algorithm for judgement of given argumentation, which can be applied to make a strategy to win.

Propagation of Action Knowledge in Multi-Agent Systems

Building on a general theory of action by Reiter and his colleagues, Bacchus et al. give an account for formalizing degrees of belief and noisy actions in the situation calculus. Unfortunately, there is no clear solution to the projection problem for the formalism. And, while the model has epistemic features, it is not obvious what the agent's knowledge base should look like. Also, reasoning about uncertainty essentially resorts to second-order logic. In recent work, Gabaldon and Lakemeyer remedy these shortcomings somewhat, but here too the utility seems to be restricted to queries (with action operators) about the initial theory. In this paper, we propose a fresh amalgamation of a modal fragment of the situation calculus and uncertainty, where the idea will be to update the initial knowledge base, containing both ordinary and (certain kinds of) probabilistic beliefs, when noisy actions are performed. We show that the new semantics has the right properties, and study a special case where updating probabilistic beliefs is computable. Our ideas are closely related to the Lin and Reiter notion of progression.

The Wumpus World problem is a classic AI challenge in which an agent must navigate a maze-like environment to find and retrieve gold while avoiding obstacles such as pits and a monster called the Wumpus. This project aims to implement a logical agent to solve the Wumpus World problem using a combination of first-order logic and the Minimax algorithm. The environment is represented as a set of logical propositions. The agent's knowledge base is expressed as a set of first-order logic rules describing the relationships between the propositions. The Minimax algorithm is used to generate a game tree and compute the minimax values of the nodes in the tree, allowing the agent to make informed decisions about which actions to take. The implementation demonstrates the versatility and expressiveness of logical agents for solving complex AI problems and highlights the potential benefits of using first-order logic in AI systems.

In [4] a new epistemic logic \(\textsf {LDA}\) of explicit and implicit belief was introduced, and in [5] we presented a tableau-based satisfability checking procedure as well as a dynamic extension for it. Based on such procedure, we created a portable software implementation that works for the family of multi-agent epistemic logics, as well as for the proposed dynamic extension. This software implementation runs as a library for the most common operative systems, also runs in popular IoT and robot hardware, as well as cloud environments and in server-less configurations.

In this paper we propose an explicit form of knowledge-based programming. Our initial motivation is the distributed implementation of game-theoretical algorithms, but we abstract away from the game-theoretical details and describe a general scenario, where a group of agents each have some initially private bits of information which they can then communicate to each other. We draw on existing literature to give a formal model using modal logic to represent the knowledge of the agents as well as how that knowledge changes as they communicate. We sketch an implementation which enables processes in a distributed system to explicitly evaluate knowledge formulae. Then we prove that the implementation captures the formal model, and therefore correctly reflects the general scenario. Finally we look at how our approach lends itself to generalisations, and discuss application perspectives.

The idealizations resulting from the use of Kripke semantics in Epistemic Logic are inherited by formalizations of group epistemic notions. For example, distributed knowledge is often taken to reflect the potential knowledge of a group: what agents would know if they had unbounded means of communication and deductive ability. However, this does not specify whether/how this potential can be actualized, especially since real people are not unbounded reasoners. Inspired by experiments on group reasoning, we identify two dimensions of actualizing distributed knowledge: communication and inference. We build a dynamic framework with effortful actions accounting for both, combining impossible-worlds semantics and action models, and we provide a method for extracting a sound and complete axiomatization.

Text mining (TM) and computational linguistics (CL) are computationally intensive fields where many tools are becoming available to study large text corpora and exploit the use of corpora for various purposes. In this chapter we will address the problem of building conversational agents or chatbots from corpora for domain-specific educational purposes. After addressing some linguistic issues relevant to the development of chatbot tools from corpora, a methodology to systematically analyze large text corpora about a limited knowledge domain will be presented. Given the Artificial Intelligence Markup Language as the “assembly language” for the artificial intelligence conversational agents we present a way of using text corpora as seed from which a set of “source files” can be derived. More specifically we will illustrate how to use corpus data to extract relevant keywords, multiword expressions, glossary building and text patterns in order to build an AIML knowledge base that could be later used to build interactive conversational systems. The approach we propose does not require deep understanding techniques for the analysis of text. As a case study it will be shown how to build the knowledge base of an English conversational agent for educational purpose from a child story that can answer question about characters, facts and episodes of the story. A discussion of the main linguistic and methodological issues and further improvements is offered in the final part of the chapter.

Purpose Human–agent interaction (HAI) is increasingly influencing our personal and work lives through the proliferation of conversational agents (CAs) in various domains. As such, these agents combine intuitive natural language interactions by also delivering personalization through artificial intelligence capabilities. However, research on CAs as well as practical failures indicates that CA interaction oftentimes fails miserably. To reduce these failures, this paper introduces the concept of building common ground for more successful HAIs. Design/methodology/approach Based on a systematic literature analysis, we identified 38 articles meeting the eligibility criteria. We critically reviewed this body of knowledge within a formal narrative synthesis structured around the use of common ground in the interaction with CAs. Findings Based on the systematic review, our analysis reveals five mechanisms for achieving common ground: embodiment, social features, joint action, knowledge base and mental model of conversational agent. We point out the relationships between these mechanisms as they are related to each other in directional and bidirectional ways. Research limitations/implications Our findings contribute to theory with several implications for CA research. First, we provide implications about the organization of common ground mechanisms for CAs. Second, we provide insights into the mechanisms and nomological network for achieving common ground when interacting with CAs. Third, we provide a broad research agenda for future CA research that centers around the important topic of common ground for HAI. Originality/value We offer novel insights into grounding mechanisms and highlight the potentials when considering common ground in different HAI processes. Consequently, we secure further understanding and deeper insights of possible mechanisms of common ground to shape future HAI processes.

Parametric logic is a framework that generalises classical first-order logic. A generalised notion of logical consequence—a form of preferential entailment based on a closed world assumption—is defined as a function of some parameters. A concept of possible knowledge base—the counterpart to the consistent theories of first-order logic—is introduced. The notion of compactness is weakened. The degree of weakening is quantified by a nonnull ordinal—the larger the ordinal, the more significant the weakening. For every possible knowledge base T, a hierarchy of sentences that are generalised logical consequences of T is built. The first layer of the hierarchies corresponds to sentences that can be obtained by a deductive inference, characterised by the compactness property. The second layer of the hierarchies corresponds to sentences that can be obtained by an inductive inference, characterised by the property of weak compactness quantified by 1. Weaker forms of compactness—quantified by nonnull ordinals—determine higher layers in the hierarchies, corresponding to more complex inferences. The naturalness of the hierarchies built over the possible knowledge bases is attested by fundamental connections with notions from Learning theory and from topology. The naturalness of the hierarchies built over the possible knowledge bases is attested by fundamental connections with notions from Learning theory—classification in the limit, with or without a bounded number of mind changes—and from topology—in reference to the Borel and the difference hierarchies. In this paper, we introduce the key model-theoretic aspects of Parametric logic, justify the concept of the knowledge base, define the hierarchies of generalised logical consequences and illustrate their relevance to Nonmonotonic reasoning. More specifically, we show that the degree of nonmonotonicity that is required to infer a sentence can be characterised by the least nonnull ordinal that quantifies the weakening of compactness used to locate the inferred sentence in the hierarchies.

This paper is concerned with a possible mechanism for learning the meanings of quantifiers in natural language. The meaning of a natural language construction is identified with a procedure for recognizing its extension. Therefore, acquisition of natural language quantifiers is supposed to consist in collecting procedures for computing their denotations. A method for encoding classes of finite models corresponding to given quantifiers is shown. The class of finite models is represented by appropriate languages. Some facts describing dependencies between classes of quantifiers and classes of devices are presented. In the second part of the paper examples of syntax-learning models are shown. According to these models new results in quantifier learning are presented. Finally, the question of the adequacy of syntax-learning tools for describing the process of semantic learning is stated.