In this paper we address the problem of generating preferred plans by combining the procedural control knowledge specified by Hierarchical Task Networks (HTNs) with rich user preferences. To this end, we extend the popular Planning Domain Definition Language, PDDL3, to support specification of simple and temporally extended preferences over HTN constructs. To compute preferred HTN plans, we propose a branch-and-bound algorithm, together with a set of heuristics that, leveraging HTN structure, measure progress towards satisfaction of preferences. Our preference-based planner, HTNPLAN-P, is implemented as an extension of the SHOP2 planner. We compared our planner with SGPLAN5 and HPLAN-P — the top performers in the 2006 International Planning Competition preference tracks. HTNPLAN-P generated plans that in all but a few cases equalled or exceeded the quality of plans returned by HPLAN-P and SGPLAN5. While our implementation builds on SHOP2, the language and techniques proposed here are relevant to a broad range of HTN planners.

Runtime commitment verification is an important, open issue in multiagent research. To address it, we build on Yolum and Singh's formalization of commitment operations, on Chittaro and Montanari's cached event calculus, and on the SCIFF abductive logic programming proof-procedure. We propose a framework consisting of a declarative and compact language to express the domain knowledge, and a reactive and complete procedure to track the status of commitments effectively, producing provably sound and irrevocable answers.

Dialogue-based Question Answering (QA) is a highly complex task that brings together a QA system including various natural language processing components (i.e., components for question classification, information extraction, and retrieval) with dialogue systems for effective and natural communication. The dialogue-based access is difficult to establish when the QA system in use is complex and combines many different answer services with different quality and access characteristics. For example, some questions are processed by opendomain QA services with a broad coverage. Others should be processed by using a domain-specific instance ontology for more reliable answers. Different answer services may change their characteristics over time and the dialogue reaction models have to be updated according to that. To solve this problem, we developed introspective methods to integrate adaptable models of the answer services. We evaluated the impact of the learned models on the dialogue performance, i.e., whether the adaptable models can be used for a more convenient dialogue formulation process. We show significant effectiveness improvements in the resulting dialogues when using the machine learning (ML) models. Examples are provided in the context of the generation of system-initiative feedback to user questions and answers, as provided by heterogeneous information services.

There has been intense interest in hierarchical reinforcement learning as a way to make Markov decision process planning more tractable, but there has been relatively little work on autonomously learning the hierarchy, especially in continuous domains. In this paper we present a method for learning a hierarchy of actions in a continuous environment. Our approach is to learn a qualitative representation of the continuous environment and then to define actions to reach qualitative states. Our method learns one or more options to perform each action. Each option is learned by first learning a dynamic Bayesian network (DBN). We approach this problem from a developmental robotics perspective. The agent receives no extrinsic reward and has no external direction for what to learn. We evaluate our work using a simulation with realistic physics that consists of a robot playing with blocks at a table.

We study the computational complexity of conjunctive query answering w.r.t. ontologies formulated in fragments of the description logic SHIQ. Our main result is the identification of two new sources of complexity: the combination of transitive roles and role hierarchies which results in 2ExpTime-hardness, and transitive roles alone which result in coNExpTime-hardness. These bounds complement the existing result that inverse roles make query answering in SHIQ 2ExpTime-hard.  We also show that conjunctive query answering with transitive roles, but without inverse roles and role hierarchies, remains in ExpTime if the ABox is tree-shaped.

This paper studies the solving of finite-domain action planning problems with discrete action costs and soft constraints. For sequential optimal planning, a symbolic perimeter database heuristic is addressed in a bucket implementation of A*. For computing net-benefits, we propose symbolic branch-and-bound search together with some search refinements. The net-benefit we optimize is the total benefit of satisfying the goals, minus the total action cost to achieve them. This results in an objective function to be minimized that is a linear expression over the violation of the preferences added to the action cost total.

A folksonomy refers to a collection of user-defined tags with which users describe contents published  on the Web. With the flourish of Web 2.0, folksonomies have become an important mean to develop the Semantic Web. Because tags in folksonomies are authored freely, there is a need to understand the structure and semantics of these tags in various applications. In this paper, we propose a learning approach to create an ontology that captures the hierarchical semantic structure of folksonomies. Our experimental results on two different genres of real world data sets show that our method can effectively learn the ontology structure from the folksonomies.

Kernel methods have been applied successfully in many applications. The kernel matrix plays an important role in kernel-based learning methods, but the ideal kernel matrix is usually unknown in practice and needs to be estimated. In this paper, we propose to directly learn the ideal kernel matrix (called the optimal neighborhood kernel matrix) from a pre-specified kernel matrix for improved classification performance. We assume that the pre-specified kernel matrix generated from the specific application is a noisy observation of the ideal one. The resulting optimal neighborhood kernel matrix is shown to be the summation of the pre-specified kernel matrix and a rank-one matrix. We formulate the problem of learning the optimal neighborhood kernel as a constrained quartic problem, and propose to solve it using two methods: level method and constrained gradient descent. Empirical results on several benchmark data sets demonstrate the efficiency and effectiveness of the proposed algorithms.

In Bayesian networks, a most probable explanation (MPE) is a most likely instantiation of all network variables given a piece of evidence. Solving (the decision version of) an MPE query is NP-hard. Recent work proposed a branch-and-bound search algorithm that finds exact solutions to MPE queries, where bounds are computed on a relaxed network obtained by a technique known as node splitting. In this work we study the impact of variable and value ordering on such a search algorithm. We study several heuristics based on the entropies of variables and on the notion of  nogoods, and propose a new meta-heuristic that combines their strengths. Experiments indicate that search efficiency is significantly improved, allowing many hard problems to be solved for the first time.

The recognition of the goal a user is pursing when interacting with a software application is a crucial task for an interface agent as it serves as a context for making opportune interventions to provide assistance to the user. The prediction of the user goal must be fast and a goal recognizer must be able to make early predictions with few observations of the user actions. In this work we propose an approach to automatically build an intention model from a plan corpus using Variable Order Markov models. We claim that following our approach, an interface agent will be capable of accurately ranking the most probable user goals in a time linear to the number of goals modeled.