Abstract. We describe a novel approach to tackle intention recognition, by combining dynamically configurable and situation-sensitive Causal Bayes Networks plus plan generation techniques. Given some situation, such networks enable the recognizing agent to come up with the most likely intentions of the intending agent, i.e. solve one main issue of intention recognition. And, in case of having to make a quick decision, focus on the most important ones. Furthermore, the combination with plan generation provides a significant method to guide the recognition process with respect to hidden actions and unobservable effects, in order to confirm or disconfirm likely intentions. The absence of this articulation is a main drawback of the approaches using Bayes Networks solely, due to the combinatorial problem they encounter. We explore and exemplify its application, in the Elder Care context, of the ability to perform Intention Recognition and of wielding Evolution Prospection methods to help the Elder achieve its intentions. This is achieved by means of an articulate use of a Causal Bayes Network to heuristically gauge probable general intention – combined with specific generation of plans involving preferences – for checking which such intentions are plausibly being carried out in the specific situation at hand, and suggesting actions to the Elder. The overall approach is formulated within one coherent and general logic programming framework and implemented system. The paper recaps required background and illustrates the approach via an extended application example.

In the previous chapter (Tenenbaum, Griffiths, & Niyogi, this volume), we introduced a framework for thinking about the structure, function, and acquisition of intuitive theories inspired by an analogy to the research program of generative grammar in linguistics. We argued that a principal function for intuitive theories, just as for grammars for natural

We argue that human judgments about hidden causal structure can be explained as the operation of domain-general statistical inference over causal models constructed using domain knowledge. We present Bayesian models of causal induction in two previous experiments and a new study. Hypothetical causal models are generated by theories expressing two essential aspects of abstract knowledge about causal mechanisms: which causal relations are plausible, and what functional form they take.

Leading accounts of judgment under uncertainty evaluate performance within purely statistical frameworks, holding people to the standards of classical Bayesian (Tversky & Kahneman, 1974) or frequentist (Gigerenzer & Hoffrage, 1995) norms. We argue that these frameworks have limited ability to explain the success and flexibility of people's real-world judgments, and propose an alternative normative framework based on Bayesian inferences over causal models. Deviations from traditional norms of judgment, such as "base-rate neglect", may then be explained in terms of a mismatch between the statistics given to people and the causal models they intuitively construct to support probabilistic reasoning. Four experiments show that when a clear mapping can be established from given statistics to the parameters of an intuitive causal model, people are more likely to use the statistics appropriately, and that when the classical and causal Bayesian norms differ in their prescriptions, people's judgments are more consistent with causal Bayesian norms.

Abstract not found

Abstract. We explore and exemplify the application in the Elder Care context of the ability to perform Intention Recognition and of wielding Evolution Prospection methods. This is achieved by means of an articulate use of Causal Bayes Nets (for heuristically gauging probable general intentions), combined with specific generation of plans involving preferences (for checking which such intentions are plausibly being carried out in the specific situation at hand). The overall approach is formulated within one coherent and general logic programming framework and implemented system. The paper recaps required background and illustrates the approach via an extended application example.

Explanation of why things happen is one of humans ’ most important cognitive operations. In everyday life, people are continually generating explanations of why other people behave the way they do, why they get sick, why computers or cars are not working properly, and of many other puzzling occurrences. More systematically, scientists develop theories to provide general explanations of physical phenomena such as why objects fall to earth, chemical phenomena such as why elements combine, biological phenomena such as why species evolve, medical phenomena such as why organisms develop diseases, and psychological phenomena such as why people sometimes make mental errors. This chapter reviews computational models of the cognitive processes that underlie these kinds of explanations of why events happen. It is not concerned with another sense of explanation that just means clarification, as when someone explains the U. S. constitution. The focus will be on scientific explanations, but more mundane examples will occasionally be used, on the grounds that the cognitive processes for explaining why events happen are much the same in everyday life and in science, although scientific explanations tend tobe more systematic and rigorous than everyday ones. In addition to providing a concise review of previous computational models of explanation, this chapter describes a new neural network model that shows how explanations can be performed by multimodal distributed representations.

Causal knowledge enables us to predict future events, to choose the right actions to achieve our goals, and to envision what would have happened if things had been different. Thus, it allows us to reason about observations, interventions and counterfactual possibilities.

Theories typically emphasize affordances or intentions as the primary determinant of an object’s perceived function. The HIPE theory assumes that people integrate both into causal models that produce functional attributions. In these models, an object’s physical structure and an agent’s action specify an affordance jointly, constituting the immediate causes of a perceived function. The object’s design history and an agent’s goal in using it constitute distant causes. When specified fully, the immediate causes are sufficient for determining the perceived function—distant causes have no effect (the causal proximity principle). When the immediate causes are ambiguous or unknown, distant causes produce inferences about the immediate causes, thereby affecting functional attributions indirectly (the causal updating principle). Seven experiments supported HIPE’s predictions. Function is a central construct in cognitive science and cognitive neuroscience. Cognitive psychologists have shown that the categorization of an artifact depends not only on its physical properties, but also on its function (e.g., Barton & Komatsu, 1989; Keil,

Classical approaches used to learn Bayesian network structure from data have disadvantages in terms of complexity and lower accuracy of their results. However, a recent empirical study has shown that a hybrid algorithm improves sensitively accuracy and speed: it learns a skeleton with an independency test (IT) approach and constrains on the directed acyclic graphs (DAG) considered during the search-and-score phase. Subsequently, we theorize the structural constraint by introducing the concept of super-structure S, which is an undirected graph that restricts the search to networks whose skeleton is a subgraph of S. We develop a super-structure constrained optimal search (COS): its time complexity is upper bounded by O(γm n), where γm < 2 depends on the maximal degree m of S. Empirically, complexity depends on the average degree ˜m and sparse structures allow larger graphs to be calculated. Our algorithm is faster than an optimal search by several orders and even finds more accurate results when given a sound super-structure. Practically, S can be approximated by IT approaches; significance level of the tests controls its sparseness, enabling to control the trade-off between speed and accuracy. For incomplete super-structures, a greedily post-processed version (COS+) still enables to significantly outperform other heuristic searches. Keywords: subset Bayesian networks, structure learning, optimal search, super-structure, connected 1.