Skip navigation.
New Mexico State University
Zhul library.

Technical Reports

 

2001  |  2002  |  2003  |  2005  |  2007  |  2008  | 

NMSU-CS-2007-001 
Planning with Preferences Using Constraint Logic Programming
Phan Huy Tu, Tran Cao Son, and Enrico Pontelli
Abstract: We describe the development of a constraint logic programming based system, called CPP, which is capable of generating most preferred plans with respect to a user's preference and evaluate its performance.
Download document:    pdf
 
NMSU-CS-2007-002 
Approximation of Action Theories and Its Application to Conformant Planning
Phan Huy Tu, Tran Cao Son, Michael Gelfond, Ricardo A. Morales
Abstract: This paper suggests a methodology for building conformant planners, which is based on recent advances in the theory of action and change and answer set programming. The development of a planner for a given dynamic domain starts with encoding the knowledge about fluents and actions of the domain as an action theory D of some action language. Our choice in this paper is AL - an action language with dynamic and static causal laws and executability conditions. An action theory D of AL defines a transition diagram T(D) containing all the possible trajectories of the domain. A transition \in T(D) iff execution of action a in state s may move the domain to state s'. The second step in the planner development consists in finding a deterministic transition diagram T^{lp}(D) such that nodes of T^{lp}(D) are partial states of D, its arcs are labeled by actions, and a path in T^{lp}(D) from an initial partial state \delta^0 to a partial state satisfying the goal \delta^f corresponds to a conformant plan for \delta^0 and \delta^f in T( D). The transition diagram T^{lp}(D) is called an 'approximation' of T(D). We claim that a concise description of an approximation of T(D) can often be given by a logic program \pi(D) under the answer sets semantics. Moreover, complex initial situations and constraints on plans can be also expressed by logic programming rules and included in \pi(D). If this is possible then the problem of finding a parallel or sequential conformant plan can be reduced to computing answer sets of \pi(D). This can be done by general purpose answer set solvers. If plans are sequential and long then this method can be too time consuming. In this case, \pi(D) is used as a specification for a procedural graph searching conformant planning algorithm. The paper illustrates this methodology by building several conformant planners which work for domains with complex relationship between the fluents. The efficiency of the planners is experimentally evaluated on a number of new and old benchmarks. In addition we show that for a subclass of action theories of AL our planners are complete, i.e., if in T^{lp}(D) we cannot get from \delta^0 to a state satisfying the goal \delta^f then there is no conformant plan for \delta^0 and \delta^f in T({\cal D}). In general our planners are incomplete and thus can miss some plans. Fortunately this seems to be a rare occasion, and the corresponding examples are rather artificial.
Download document:    pdf
 
NMSU-CS-2007-003 
On the Completeness of Approximation Based Reasoning and Planning in Action Theories with Incomplete Information
Tran Cao Son and Phan Huy Tu
Abstract: In this paper we study the completeness of the 0-approximation with respect to the possible world semantics for action theories with incomplete information, propose a sufficient condition for its completeness, and define the notion of a {\em complete action theory}. We introduce the notion of a {\em decisive sets of fluents}, that can be used to transform an action theory into an equivalent and complete one, and develop a polynomial time algorithm for computing decisive sets. We then extend the results to the conformant planning setting and use them in the development of a sound and complete conformant planner. Finally, we compare our planner with other state-of-the-art conformant planners on benchmarks found in the literature.
Download document:    pdf
 
NMSU-CS-2007-004 
Planning for Biochemical Pathways: A Case Study of Answer Set Planning in Large Planning Problem Instances
Tran Cao Son and Enrico Pontelli
Abstract: The paper describes an experiment of answer set planning in biochemical pathway planning. The focus is on large planning problem instances. It is shown that well-known planning techniques, such as planning graph analysis, landmarks recognition, and planning using landmarks are useful in answer set planning and can be easily incorporated in an answer set planning system.
Download document:    pdf
 
NMSU-CS-2007-005 
Parallel Processing in Conformant Planning: Methodologies and Experiments
To Thanh Son, Phan Huy Tu, Enrico Pontelli, Tran Cao Son
Abstract: The paper investigates the application of parallel processing techniques to the problem of computing solutions of conformant planning problems. The work is motivated by the wider availability of small and medium scale parallel platforms (e.g., multi-core desktops and Beowulf clusters) and the growing simplicity in parallel programming. We explore the extraction of orthogonal forms of parallelism from a competitive conformant planner. The experiments demonstrate the potential for improved performance by taking advantage of the concurrency of modern computing platforms.
Download document:    pdf
 
NMSU-CS-2007-006 
Development of a Database and its Implementation to Model the Impact of Rapid Anthropogenic Land Cover Changes on Hantavirus Ecology in Paraguay
Tran Cao Son, Phan Huy Tu , Linda Allen, Yong-kyu Chu, D. Goodin, S. Hutchinson, Colleen B. Jonsson, Robert D . Owen, Enrico Pontelli
Abstract: Hantavirus pulmonary syndrome (HPS) is an emerging disease that is transmitted to humans by hantaviruses harbored and shed by rodents of the family Muridae. The driving forces behind HPS outbreaks that often correlate to a rapid expansion of virus in the rodent population prior to spill over into humans are poorly understood. We hypothesize that the ecological dynamics of the virus-host-landscape system is crucial for development of models to predict HPS outbreaks. The hypothesis is investigated using a multidisciplinary approach, combining modeling and longitudinal field studies of the hantavirus-host-landscape system in Paraguay at several spatial scales. To facilitate the modeling we have developed an Oracle-based database system to store all field and laboratory records. Initial analysis of the data showed that human disturbed landscapes are more likely to harbor seropositive rodents, but provides little insight into why disturbance should be associated with viral or antibody presence. To address the underlying causes of hantavirus dynamics within the rodent population, we developed a male/female SIR epidemic model based on information garnered from the database. These new models capture some of the realistic dynamics of the male/female rodent hantavirus interaction, including higher seroprevalence in males and variability in seroprevalence levels. This database can be employed to evaluate the relationship between land use/land cover dynamics, and rodent-hantavirus presence using remote sensing and trapping data.
Download document:   
 
NMSU-CS-2007-007 
Using and Interfacing Background Knowledge in Story Understanding
Nemecio R. Chavez, Jr., Heather D. Pfeiffer, Roger T. Hartley
Abstract: Described is two background knowledge systems used to provide basic knowledge needed for story understanding. The basic knowledge is information about the world that a 6 or 7 year old child might know. A story understanding system also provides an added layer of knowledge that will be referred to as prototypes. The complete knowledge needed for understanding comes in two forms: 1) knowl-edge about objects in the world, and 2) prototypes, which describe higher-order knowledge or experience. The story understanding system uses multi-agents to combine this knowledge into a meaningful structure from which under-standing can be demonstrated, i.e. question answering.
Download document:    pdf