Fusun Yaman

AAAI Conference 2013 Conference Paper

A Morphogenetically Assisted Design Variation Tool

• Aaron Adler
• Fusun Yaman
• Jacob Beal
• Jeffrey Cleveland
• Hala Mostafa
• Annan Mozeika

The complexity and tight integration of electromechanical systems often makes them “brittle” and hard to modify in response to changing requirements. We aim to remedy this by capturing expert knowledge as functional blueprints, an idea inspired by regulatory processes that occur in natural morphogenesis. We then apply this knowledge in an intelligent design variation tool. When a user modifies a design, our tool uses functional blueprints to modify other components in response, thereby maintaining integration and reducing the need for costly search or constraint solving. In this paper, we refine the functional blueprint concept and discuss practical issues in applying it to electromechanical systems. We then validate our approach with a case study applying our prototype tool to create variants of a miniDroid robot and by empirical evaluation of convergence dynamics of networks of functional blueprints.

AAAI Conference 2013 Conference Paper

Temporal Milestones in HTNs

• Fusun Yaman
• Brett Benyo
• Alice Mulvehill

We present temporal milestones for hierarchical task networks to enable the complex synchronization of tasks. A temporal milestone of a task is an intermediate event that occurs during the execution of a complex task, e. g. , the start time, the end time or a milestone of any of its subtasks. Unlike landmark variables, introduced in existing work, temporal milestones respect the task abstraction boundaries and preserve structural properties enabling much more efficient reasoning. Furthermore, temporal milestones are as expressive as landmark variables. We provide analytical and empirical evidence to support these claims.

AIJ Journal 2011 Journal Article

Democratic approximation of lexicographic preference models

• Fusun Yaman
• Thomas J. Walsh
• Michael L. Littman
• Marie desJardins

Lexicographic preference models (LPMs) are an intuitive representation that corresponds to many real-world preferences exhibited by human decision makers. Previous algorithms for learning LPMs produce a “best guess” LPM that is consistent with the observations. Our approach is more democratic: we do not commit to a single LPM. Instead, we approximate the target using the votes of a collection of consistent LPMs. We present two variations of this method—variable voting and model voting—and empirically show that these democratic algorithms outperform the existing methods. Versions of these democratic algorithms are presented in both the case where the preferred values of attributes are known and the case where they are unknown. We also introduce an intuitive yet powerful form of background knowledge to prune some of the possible LPMs. We demonstrate how this background knowledge can be incorporated into variable and model voting and show that doing so improves performance significantly, especially when the number of observations is small.

TIST Journal 2011 Journal Article

RECYCLE

• Karen Zita Haigh
• Fusun Yaman

A workflow is a model of a process that systematically describes patterns of activity. Workflows capture a sequence of operations, their enablement conditions, and data flow dependencies among them. It is hard to design a complete and correct workflow from scratch, while it is much easier for humans to demonstrate the solution than to state the solution declaratively. This article presents RECYCLE, our approach to learning workflow models from example demonstration traces. RECYCLE captures control flow, data flow, and enablement conditions of an underlying workflow process. Unlike prior work from workflow mining and AI planning literature, (1) RECYCLE can learn from a single demonstration trace with loops, (2) RECYCLE learns both loop and conditional branch structure, and (3) RECYCLE handles data flow among actions. In this article, we describe the phases of RECYCLE's learning algorithm: substructure analysis and node abstraction. To ground the discussion, we present a simplified flight reservation system with some of the important characteristics of the real domains we worked with. We present some results from a patient transport domain.

IJCAI Conference 2009 Conference Paper

• Fusun Yaman
• Tim Oates
• Mark Burstein

Existing work on workflow mining ignores the dataflow aspect of the problem. This is not acceptable for service-oriented applications that use Web services with typed inputs and outputs. We propose a novel algorithm WIT (Workflow Inference from Traces) which identifies the context similarities of the observed actions based on the dataflow and uses model merging techniques to generalize the control flow and the dataflow simultaneously. We identify the class of workflows that WIT can learn correctly. We implemented WIT and tested it on a real world medical scheduling domain where WIT was able to find a good approximation of the target workflow.

ICML Conference 2008 Conference Paper

Democratic approximation of lexicographic preference models

• Fusun Yaman
• Thomas J. Walsh 0001
• Michael L. Littman
• Marie desJardins

IJCAI Conference 2007 Conference Paper

• Austin Parker
• Fusun Yaman
• Dana Nau
• VS Subrahmanian

There are numerous cases where we need to reason about vehicles whose intentions and itineraries are not known in advance to us. For example, Coast Guard agents tracking boats don't always know where they are headed. Likewise, in drug enforcement applications, it is not always clear where drug-carrying airplanes (which do often show up on radar) are headed, and how legitimate planes with an approved flight manifest can avoid them. Likewise, traffic planners may want to understand how many vehicles will be on a given road at a given time. Past work on reasoning about vehicles (such as the "logic of motion" by Yaman et. al. ) only deals with vehicles whose plans are known in advance and don't capture such situations. In this paper, we develop a formal probabilistic extension of their work and show that it captures both vehicles whose itineraries are known, and those whose itineraries are not known. We show how to correctly answer certain queries against a set of statements about such vehicles. A prototype implementation shows our system to work efficiently in practice.

UAI Conference 2007 Conference Paper

More-or-Less CP-Networks

• Fusun Yaman
• Marie desJardins

Preferences play an important role in our everyday lives. CP-networks, or CP-nets in short, are graphical models for representing conditional qualitative preferences under ceteris paribus ("all else being equal") assumptions. Despite their intuitive nature and rich representation, dominance testing with CP-nets is computationally complex, even when the CP-nets are restricted to binary-valued preferences. Tractable algorithms exist for binary CP-nets, but these algorithms are incomplete for multi-valued CPnets. In this paper, we identify a class of multivalued CP-nets, which we call more-or-less CPnets, that have the same computational complexity as binary CP-nets. More-or-less CP-nets exploit the monotonicity of the attribute values and use intervals to aggregate values that induce similar preferences. We then present a search control rule for dominance testing that effectively prunes the search space while preserving completeness.

IJCAI Conference 2005 Conference Paper

A Motion Closed World Asumption

• Fusun Yaman
• Dana Nau
• V. S

Yaman et. al. [Yaman et al. , 2004] introduce “go theories” to reason about moving objects. In this paper, we show that this logic often does not allow us to infer that an object is not present at a given place or region, even though common sense would dictate that this is a reasonable inference to make. We define a class of models of go-theories called coherent models. We use this concept to define a motion closed world assumption (MCWA) and develop a notion of MCWA-entailment. We show that checking if a go-theory has a coherent model is NP-complete. An in atom checks if a given object is present in a given region sometime in a given time interval. We provide sound and complete algorithms to check if a ground in literal (positive or negative in atom) can be inferred from a gotheory using the MCWA. In our experiments our algorithms answer such queries in less than 1 second when there are up to 1, 000 go-atoms per object.

IJCAI Conference 2005 Conference Paper

Going Far, Logically

• Fusun Yaman
• Dana Nau
• V. S

There are numerous applications where we need to ensure that multiple moving objects are sufficiently far apart. Furthermore, in many moving object domains, there is positional indeterminacy — we are not 100% sure exactly when a given moving object will be at a given location. [Yaman et al. , 2004] provided a logic of motion but did not provide algorithms to ensure that moving objects are kept sufficiently far apart. In this paper, we extend their logic to include a “far” predicate. We develop the CheckFar algorithm that checks if any given two objects will always be sufficiently far apart at during a time interval. We have run a set of experiments showing that our CheckFar algorithm scales very well.

KR Conference 2004 Conference Paper

Logic of Motion

• Dana Nau
• V. S. Subrahmanian
• Fusun Yaman

There are numerous applications such as air traffic management, cellular phone location tracking, and vehicle protection systems where there is a critical need to reason about moving objects. In this paper, we propose a formal logic of motion (LOM for short). We provide a formal syntax for LOM, as well as a model theory for LOM. In addition, we develop algorithms to check consistency of LOM theories, as well as to answer certain kinds of queries posed to LOM theories. We have implemented these algorithms in a prototype LOM system - we describe experiments showing that such queries can be efficiently executed in practice.