http://rpgoldman.real-time.com/Robert P. Goldman New Stuff
New plan recognition paper from AAAI-2011 with Chris Geib.
Computer Science Research
SMITE (Computer Security, Intrusion Detection)I am currently working on the SMITE project, as part of DARPA's Scalable Network Monitoring program. SMITE aims at performing intrusion detection on ultra high speed network traffic (up to 100 Gbps).
My particular role in the project is to work on what is called intrusion detection system (IDS) "correlation" (correlation is actually a misleading term, what is really meant here is "fusion.")
Even more specifically, my work is in the component that is called the "evidence assessor." This component uses qualitative Bayesian reasoning to weigh competing hypotheses for clusters of IDS resports. Often these competing explanations weigh malign (malware sweeping the network in order to spread) against benign (print daemons looking for networked printers) explanations; the evidence assessor helps us address the problem of false positives that has plagued IDSes.
This project continues work on techniques that were first developed in the ARGUS/Scyllarus project.
Papers from SMITE:
"Model-based Intrusion Assessment in Common Lisp ," with Steven A. Harp, in working notes of the ICAPS 2009 Workshop on Computer Security and Artificial Intelligence (SecArt), Thessaloniki, Greece, September, 2009. This paper is a mild revision and reformatting of the ILC paper, listed below, which is the authoritative (archival) version.
"Model-based Intrusion Assessment in Common Lisp," with Steven A. Harp Proceedings of the International Lisp Conference, Cambridge, MA, 2009.
Integrated learning (Planning, Knowledge Representation, Semantic Web)
My most recently concluded activity was working on the BBN-led POIROT team for DARPA's Integrated Learning program, which aimed at integrating together a large number of intelligent systems to learn workflows from observing humans using web services.
Papers from Integrated Learning:
2010
"Shopper: a system for executing and simulating expressive plans," Proceedings of ICAPS 2010, Toronto, Canada, 2010. short paper with John Maraist. There was also be a demonstration of this system at ICAPS 2010.
2009
"A Semantics for HTN Methods," in the Proceedings of ICAPS 2009, Thessaloniki, Greece, September, 2009.
"Partial Observability, Quantification, and Iteration for Planning: Work in Progress," in working notes of the ICAPS 2009 Workshop on Generalized Planning, Thessaloniki, Greece, September, 2009.
"SHOPPER: Interpreter for a high-level web services language," with John Maraist, in the Proceedings of the International Lisp Conference, Cambridge, MA, March 2009.
"LTML -- a language for representing semantic web service workflow procedures," with Mark Burstein, et al., To appear in the working notes of the "Semantics for the Rest of Us" workshop at the 8th International Semantic Web Conference (ISWC 2009).
2008
"Using Classical Planners to Solve Nondeterministic Planning Problems," with Ugur Kuter, Dana Nau and Elnatan Reisner, in the Proceedings of ICAPS 2008, Sydney, Australia, September 2008.
2007
"HOTRiDE: Hierarchical Ordered Task Replanning in Dynamic Environments," with N. Fazil Ayan, Ugur Kuter, and Fusun Yaman, on replanning and execution monitoring of HTNs.
"Conditionalization: Adapting forward-chaining planners to partially observable environments," with Ugur Kuter, Dana Nau, Elnatan Reisner, Proceedings of ICAPS-2007.
Planning for Uninhabited Air Vehicles (UAVs)
I've been doing a lot of work on planning for UAVs in the past years. This is closely related to my interest in Planning for Intelligent UIs, since we're trying to use AI planners to make it easier to command complex behaviors of autonomous systems. Unfortunately, the pace of this work has been such that I have not had time to write much about it. Here are some snippets:
With Mark Boddy, I taught a tutorial on planner domain modeling at ICAPS05 (June 2005).
Most of this work has been funded under two U.S. government SBIR contracts: PVACS and TUSC.
At the ICAPS 2004 workshop on Connecting Planning Theory with Practice, I presented a paper with some lessons learned about using the SHOP2 planner in a UAV control interface.
I've been contributing to the maintenance of the SHOP2 HTN planning system.
Planning and Controller Synthesis
My work in automatic controller synthesis has mostly been done in collaboration with my colleague, David Musliner. He introduced me to the CIRCA architecture he developed (while at the University of Michigan). CIRCA is a novel architecture for doing intelligent control in hard real-time. I have since done work on incorporating model-checking into the architecture, doing abstraction-based controller synthesis, and extending the modeling capabilities for hierarchical control.
In 2001 I participated in the NASA New Millennium Program planning process, and to be on the program committee of the AAAI Symposium on Model-Based Validation of Intelligence, to be held at Stanford in March, 2001.
In 2002, I presented results from this work at the 2002 Workshop on Hybrid Control: Control and Computation (HSCC-2002).
In 2005, Musliner, Pelican, and I wrote a paper about incremental verification in controller synthesis at the Third Workshop on Model Checking and Artificial Intelligence (MoChArt05).
February 2006: Our paper on incremental verification in controller synthesis has been published in Electronic Notes in Theoretical Computer Science. You may find the paper through Elsevier's ScienceDirect website.
For recent papers in this area, click here.
My work on planning and controller synthesis grew out of earlier work on planning in conditions of incomplete information.
Information Fusion
I am interested in the use of Bayesian networks (belief networks, causal probabilistic networks) to do information fusion. I have worked on this in the context of troubleshooting abnormal situations in oil refineries (in the context of Honeywell's project on Abnormal Situation Management), and am now exploring the use of belief networks for information fusion in computer network intrusion detection (see the Argus project).
Planning with Incomplete Information
I became interested in work on planning under uncertainty as an outgrowth of my interest in general aspects of reasoning under uncertainy. Some of my early work in this area was concerned with developing a better understanding of algorithms for conditional planning, such as CNLP and Cassandra. This led me to work on formalizing the problem of conditional planning, and the development of algorithms for conditional linear planning and epsilon-safe planning. Some relevant papers include:
Representing Uncertainy in Simple Planners (from KR-94): Postscript only;
Conditional Linear Planning (from AIPS-94): Postscript and pdf;
Epsilon-Safe Planning (from UAI-94): Postscript and pdf;
Expressive Planning and Explicit Knowledge (from AIPS-96): Postscript and pdf;
These papers were all written with Mark S. Boddy at Honeywell Laboratories.
Planning for Intelligent User Interfaces
Through work with colleagues in user-centered design, notably Chris Miller, I have become interested in exploring the ways that planning systems, particularly constraint-based planning systems, can provide a user-interface to advanced control systems. One of the systems my group at Honeywell built was a multi-agent constraint-based planner (see below). Much of that work was done in conjunction with my colleague Karen Haigh.
In the course of this work, I have become interested in using optimization as a way to manage user interaction for intelligent systems. With some colleagues, I have written this work up in a paper that appeard in the 2005 proceedings of the American Helicopter Society's International Specialists' Meeting on Unmanned Rotorcraft: "Optimizing to Satisfice."
In 2006, a paper entitled "Delegation Interfaces for a Dynamic and Unpredictable Task: A Mobile Target Tracking Example," with my colleagues Chris Miller and Harry Funk, was presented at the NATO RTO Human Factors and Medicine Panel (HFM) Symposium which was held in Biarritz, France, 9-11 October 2006. I will make that paper available here if that's legal.
A paper about the general approach, but written for a human factors audience, rather than an AI audience appeared at the HFES conference: "The Playbook Approach to Adaptive Automation," Proceedings of the Human Factors and Ergonomics Society 49th Annual Meeting, 2005.
Miscellaneous
From 2005 through 2007 I worked on the DARPA Coordinators program. This was a research program investigating coordination across multiple agents that execute complex, hierarchical tasks, in a dynamic environment. See multi-agent coordination.
My earliest work at Honeywell Technology Center was done on planning under uncertainty, and on constraint-based scheduling, specifically for batch manufacturing (a kind of manufacturing that is a hybrid between continuous and discrete manufacturing. For information on this work, and an off-print of an article from IEEE Intelligent Systems, send me some email.
Another offshoot of my work on controller synthesis has been an interest in verification. Here is a paper I co-authored, appearing in the proceedings of the 2001 Spin Workshop, describing work on verification at Honeywell Laboratories.
This web page is currently very much under construction.
Curriculum Vitae
My CV is available as HTML for easier browsing.
I have removed the MS Word version because keeping it up-to-date was difficult, especially in the face of revisions to the software and file formats. If you must have it in MS Word, please try to import the HTML and contact me if you have any problems.
I will also make publications available through this web site.
Recent papers
Here are some of my recent papers in various interest areas:
Planning: There are a number of papers that should be added here. See also CIRCA papers and Constraint-based planning, and Integrated Learning.
"Durative Planning in HTNs," short paper in Proceedings of ICAPS 2006, June 2006.
Multi-agent coordination
Coordinating Highly Contingent Plans: Biasing Distributed MDPs Towards Cooperative Behavior, with David J. Musliner, Ed Durfee and Mark Boddy, paper for the ICAPS 2008 Multiagent Planning Workshop.
"Flexibly Integrating Deliberation and Execution in Decision-Theoretic Agents," with David Musliner, Jim Carciofini, Ed Durfee, Jianhui Wu, and Mark Boddy. Paper for the ICAPS-2007 workshop on Planning and Plan Execution for Real-World Systems: Principles and Practices for Planning in Execution.
"Unrolling" Complex Task Models into MDPs, paper for the working notes of the AAAI 2007 Spring Symposium on Game Theory and Decision Theory. Describes work on compiling complex hierarchical task models (specifically in the TAEMS language) into Markov Decision Processes. Joint work with Dave Musliner, Mark Boddy, Ed Durfee, and Jianhui Wu.
AAAI Spring Symposium 2006 paper, Coordinated Plan Management Using Multiagent MDPs, describes work I'm doing with researchers from Honeywell, Adventium Labs, and the University of Michigan on the DARPA Coordinators program.
"CTAEMS Language Specification," with Mark Boddy, Bryan Horling, John Phelps, Regis Vincent, A. Chris Long, Bob Kohout, and Rajiv Maheswaran. This was the specification for a dialect of the TAEMS multi-agent HTN notation, developed for the DARPA Coordinators program that clarified a number of issues in the original TAEMS language design.
Planning and AI Planning: I had a paper, with Jeff Abbott and John R. Surdu, on a simulation-based paradigm for military operations planning in the September 2008, SIW workshop: "Task-based Approach to Planning".
Plan recognition:
"Recognizing Plans with Loops Represented in a Lexicalized Grammar," with Christopher W. Geib, in the Proceedings of the AAAI Conference, San Francisco, CA, August 2011.
I was one of the chairs of a Dagstuhl workshop on plan recognition, held in April, 2011. The final report, which I edited with the other chairs (Chris Geib, Henry Kautz and Tamim Asfour) is available here.
"Handling Looping and Optional Actions in YAPPR," with Christopher W. Geib, in the AAAI Workshop on Plan, Activity and Intent Recognition (PAIR), Atlanta, GA, 2010.
"Plan Libraries for Plan Recognition: Do We Really Know What They Model?", with Froduald Kabanza and Philipe Bellefeuille, in the AAAI Workshop on Plan, Activity and Intent Recognition (PAIR), Atlanta, GA, 2010.
"A Probabilistic plan recognition algorithm based on plan tree grammars," with Christopher W. Geib, Artificial Intelligence, volume 117, number 11, 2009.
"A New Probabilistic Plan Recognition Algorithm Based on String Rewriting," joint work with Chris Geib and John Maraist, in the Proceedings of ICAPS-2008, Sydney, Australia, 2008.
A somewhat more updated, and more conjectural version of the above paper, for presentation at the Plan, Activity and Intent Recognition Workshop at IJCAI-2009 is here.
"Partial Observability and Probabilistic Plan/Goal Recognition," joint work with Chris Geib, in the proceedings of the 2005 Workshop on Modeling Others from Observations.
A New Theory of Plan Recognition, published in the Proceedings of the 1999 Conference on Uncertainty in Artificial Intelligence, joint work with Chris Geib and Chris Miller. Available in pdf or postscript. (For some reason, at least on my screen, the pdf looks horrible, and my version of gnu ghostview can't open the postscript, although kghostview can. Not sure why. Sorry.)
Reactive Planning/Controller Synthesis: I have been working, in the context of the CIRCA intelligent control architecture, on the problem of automatically synthesizing hard real-time controllers. Some recent papers in this area are:
The Evolution of CIRCA, a Theory-Based AI Architecture with Real-Time Performance Guarantees , a AAAI Spring Symposium paper from 2008 giving an overview of the CIRCA architecture and its history.
Incremental Verification for On-the-Fly Controller Synthesis, with David Musliner and Michael Pelican, at MoChart05.
For the AAAI 2000 Symposium on Hybrid Control; you can find it here.
For the 2002 Workshop on Hybrid Systems: Computation and Control. Click here for my paper.
A similar paper, with a couple of improvements, and written more for an AI audience, will be presented at the 2002 AIPS workshop on Planning via Model Checking.
"Modifying HyTech to automatically synthesize hybrid controllers," with R. Deshpande, D.J. Musliner, Jorge Tierno and Steve Pratt, Proceedings of the Conference on Decision and Control, pp. 1223-1228, Orlando, FL, 2001
Deliberation Scheduling/Metareasoning: The CIRCA architecture relies on having a real-time subsystem that provides hard real-time reactions, while its AI subsystem has the freedom to use planning algorithms that may be intractable. Recently, we have been working to get the inference of the AI subsystem under timing discipline, as well. To do this, we have been pursuing methods of deliberation scheduling, specially tailored to the CIRCA framework. Our earliest experiments in this line of work have been written-up in a paper entitled Managing Online Self-Adaptation in Real-Time Environments, available in postscript or pdf. I presented this paper at the International Workshop on Self-Adaptive Software, in mid-May 2001.
Constraint-based planning:
"MACBeth: A Multi-Agent Constraint-Based Planner" --- This paper was presented at the AAAI-2000 workshop on Constraints and AI Planning. Postscript or Adobe Acrobat.
More recently, I was co-author on a survey of Constraint-based planning that grew out of the AAAI-2000 workshop. A preprint is available in postscript. The paper will appear in AI magazine.
Computer security and information fusion:
In 2010, I was co-chair (with Mark S. Boddy and Stefan Edelkamp) of the SecArt workshop on Security and Artificial Intelligence, at AAAI in Atlanta, Georgia.
"Model-based Intrusion Assessment in Common Lisp," with Steven A. Harp, presented at the International Lisp Conference, March 2009.
Information Modeling for Intrusion Report Aggregation, Postscript and Adobe Acrobat.
Plan Recognition in Intrusion Detection Systems, Postscript and Adobe Acrobat.
A Stochastic Model for Intrusions, Adobe Acrobat presented at the 2002 Symposium on Recent Advances in Intrusion Detection (RAID-2002).
Common Lisp Build System: I have recently spent some time helping Faré (François-René Rideau) develop the new version of ASDF (Another System Definition Facility), the de facto standard build system for Common Lisp. I am proud to say that our ASDF 2 is now distributed with all major CL implementations. The project raised a number of interesting issues arising from Lisp's rich notion of a run-time environment (meaning that simply aping make is a non-starter); the way this rich environment requires us to ``hot patch'' an existing ASDF in a CL environment; the fact that there are multiple CL implementations, not just a single reference implementation; etc. We describe this work in a paper published at the 2010 International Lisp Conference:
"Evolving ASDF: More cooperation, less coordination," François-René Rideau and Robert P. Goldman, Proceedings of the International Lisp Conference, Reno, NV, 2010.
The ARGUS project
One of my last research projects at Honeywell, and one of those of which I was most proud was the ARGUS project. The ARGUS project was work in the area of computer security, specifically intrusion detection. The objective of the ARGUS project was to provide a framework for fusing the reports of multiple intrusion detection systems into a single unified view of a computer installation's security situation. This work builds on my interest in qualitative probability and Bayesian approaches to information fusion.
A couple of papers on the Argus project, were published at the DISCEX-2001 conference. Drafts are available:
Information Modeling for Intrusion Report Aggregation, Postscript and Adobe Acrobat.
Plan Recognition in Intrusion Detection Systems, Postscript and Adobe Acrobat.