Figure 2 - uploaded by Christopher W. Geib
Content may be subject to copyright.
Source publication
Previous work on the YAPPR plan recognition system pro-vided algorithms for translating conventional HTN plan li-braries into lexicalized grammars and treated the problem of plan recognition as one of parsing. To produce these gram-mars required a fixed bound for any loops within the grammar and a presented a problem for optional actions within HTN...
Similar publications
Graph Grammar (GG) is an appropriate formal language for specifying complex systems. In a GG the system states are represented by graphs and the changes between the states are described by rules. The use of GGs is interesting as there are several techniques for the specification and verification of systems that are described in this language. Besid...
Idioms are taking a vital part in effective communication as well as a crucial part of cultural inheritance. It represents the group of words together have the meaning which is different from an individual word meaning, for this metaphorical behavior idioms arise difficulties in the general machine translation system. In this paper, we have propose...
The article describes a string recognition approach, engraved in the parsers generated by Tunnel Grammar Studio that use the tunnel parsing algorithm, of how a lexer and a parser can operate on the input during its recognition. Proposed is an addition of the augmented Backus-Naur form syntax that enables the formal language to be expressed with a p...
Machine Translation is an essential approach for localization, and is especially appropriate in a linguistically diversenation like India. Automatic translation between languages which are morphologically rich and syntactically different is generally regarded as a complex task. A number of machine translation systems have been proposed in literatur...
Idioms are taking a vital part in effective communication as well as a crucial part of cultural inheritance. It represents the group of words together have the meaning which is different from an individual word meaning, for this metaphorical behavior idioms arise difficulties in the general machine translation system. In this paper, we have propose...
Citations
... Empirical evaluations show that YAPPR always outperforms PHATT, thus demonstrating the interest of the method. Further extension of YAPPR has addressed the issues of optional actions and loops in the plan library [Geib and Goldman, 2010]. Kabanza et al. introduced a heuristic weighted model counting algorithm based on YAPPR which does not Chapter 2. Recognizing the Intention of an Adversary necessitate an exhaustive enumeration of plan execution models [Kabanza et al., 2013] . ...
In this thesis, we address the problem of threat assessment, a high-level information fusion task whose main objective is to assist a decision maker in achieving a proper level of situation awareness so as to make effective and proactive decisions in possibly hostile, dynamic, and uncertain environments. Threat assessment is the problem of predicting intentional threat events and therefore it can be seen as a specific aspect of the problems of adversarial intention recognition and behavior prediction. Threat assessment is needed in adversarial situations, where several agents are competing to achieve conflicting goals in a shared environment. Adversarial situations can be modeled and analyzed using game theory which provides a formal framework for studying strategic interactions between rational decision makers. The main contribution of this thesis consists of a generic framework for threat assessment and decision support called TARGET (Threat Assessment and Response using GamE-Theory). We adopt a generative approach, thus providing an alternative to classical approaches to threat assessment based on threatening situations templates designed by subject-matter experts. Our approach combines a game-theoretic model of adversarial behavior and an inverse-planning-based approach to adversarial intention recognition. Intention recognition as inverse planning enables to infer the underlying intent driving the observed agent’s behavior by inverting a model of its planning process. The approach proposed in this thesis consists of modeling the adversary as a boundedly-rational agent whose behavior results from probabilistic planning in a stochastic game played on an attack graph. In contrast to most of the existing methods for intention recognition, our approach does not need a predefined plan library specifying the expected behaviors of the adversary. Adversarial intention recognition
is characterized by the hostility of the observed agent to the recognition process. To address this issue, we propose a set of techniques that make our system robust to deception and concealment.
The potential of our approach for intention recognition, threat assessment, and friendly decision support is evaluated on an illustrative scenario involving the protection of critical facilities.
