Showing papers by "Moshe Y. Vardi published in 2022"

Efficiency vs. Resilience: Lessons from COVID-19

Abstract: Why was the world not ready for COVID-19, in spite of many warnings over the past 20 years of the high likelihood of a global pandemic? This chapter argues that the economic goal of efficiency, focused on short-term optimization, has distracted us from resilience, which is focused on long-term optimization. Computing also seems to have generally emphasized efficiency at the expense of resilience. But computing has discovered that resilience is enabled by redundancy and distributivity. These principles should be adopted by society in the “after-COVID” era.

Quantum-Inspired Perfect Matching under Vertex-Color Constraints

Abstract: We propose and study the graph-theoretical problem EXISTS-PMVC: the existence of perfect matching under vertex-color constraints on graphs with bi-colored edges. EXISTS-PMVC is of special interest because of its motivation from quantum-state identification and quantum-experiment design, as well as its rich expressiveness, i.e., EXISTS-PMVC naturally subsumes important constrained matching problems, such as exact perfect matching. We give complexity and algorithmic results for EXISTS-PMVC under two types of vertex color constraints: (1) decision-diagram constraints (EXISTS-PMVC-DD) and (2) symmetric constraints (EXISTS-PMVC-Sym). For EXISTS-PMVC-DD, we reveal its NP-hardness by a graph-gadget technique. We prove that EXISTS-PMVC-Sym with a bounded number of colors (EXISTS-PMVC-Sym-Bounded) is polynomially equivalent with Exact Perfect Matching (XPM), which implies that EXISTS-PMVC-Sym-Bounded is in RNC on general graphs and PTIME on planar graphs. Directly applying algorithms for XPM to solve EXISTS-PMVC-Sym-Bounded is, however, impractical. We propose algorithms that natively handle EXISTS-PMVC-Sym-Bounded with considerably better complexity. Our novel results for EXISTS-PMVC provide insights into both constrained matching and scalable quantum experiment design.

ACM, ethics, and corporate behavior

Abstract: E V E R YON E IN COMPUTING is promoting ethics these days. The Vatican has issued the Rome Call for AI Ethics, which has been endorsed by many organizations, including tech companies. Facebook (now Meta) has donated millions of U.S. dollars to establish a new Institute for Ethics in Artificial Intelligence at the Technical University of Munich, since “ensuring the responsible and thoughtful use of AI is foundational to everything we do.”a Google announced it “is committed to making progress in the responsible development of AI.”b And last, but not least, ACM now requires nominators and endorsers of ACM award candidates attest that “To the best of my knowledge, the candidate ... has not committed any action that violates the ACM Code of Ethics and ACM’s Core Values.” But AI technology is the fundamental technology that underlies “Surveillance Capitalism,” defined as an economic system centered on the commodification of personal data with the core purpose of profit-making. Under the mantra of “Information wants to be free,” several tech companies have turned themselves into advertising companies. They have also perfected the technology of micro-targeted advertising, which matches ads with individual preferences. In Silicon Valley lingo, this business model is described as, “If you’re not paying for it, you’re the product.” Shoshana Zuboff arguedc eloquently about the societal risk posed by surveillance capitalism. “We can have democracy,” she wrote, “or we can have a surveillance society, but we cannot have both.” Internet companies have mastered the art of harvesting the grains of information we share with

DPMS: An ADD-Based Symbolic Approach for Generalized MaxSAT Solving

Abstract: Boolean MaxSAT, as well as generalized formulations such as Min-MaxSAT and Max-hybrid-SAT, are fundamental optimization problems in Boolean reasoning. Existing methods for MaxSAT have been successful in solving benchmarks in CNF format. They lack, however, the ability to handle 1) (non-CNF) hybrid constraints, such as XORs and 2) generalized MaxSAT problems natively. To address this issue, we propose a novel dynamic-programming approach for solving generalized MaxSAT problems with hybrid constraints -- called \emph{Dynamic-Programming-MaxSAT} or DPMS for short -- based on Algebraic Decision Diagrams (ADDs). With the power of ADDs and the (graded) project-join-tree builder, our versatile framework admits many generalizations of CNF-MaxSAT, such as MaxSAT, Min-MaxSAT, and MinSAT with hybrid constraints. Moreover, DPMS scales provably well on instances with low width. Empirical results indicate that DPMS is able to solve certain problems quickly, where other algorithms based on various techniques all fail. Hence, DPMS is a promising framework and opens a new line of research that invites more investigation in the future.

Accountability and liability in computing

Abstract: column Free Access Share on Accountability and liability in computing Author: Moshe Y. Vardi Rice University, Houston, TX Rice University, Houston, TXView Profile Authors Info & Claims Communications of the ACMVolume 65Issue 11November 2022 pp 5https://doi.org/10.1145/3565237Published:20 October 2022Publication History 0citation4,704DownloadsMetricsTotal Citations0Total Downloads4,704Last 12 Months4,704Last 6 weeks4,618 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteView all FormatsPDF

War and tech (and ACM)

Abstract: No abstract available.

Ising Model Partition Function Computation as a Weighted Counting Problem

Abstract: While the Ising model remains essential to understand physical phenomena, its natural connection to combinatorial reasoning makes it also one of the best models to probe complex systems in science and engineering. We bring a computational lens to the study of Ising models, where our computer-science perspective is two-fold: On the one hand, we consider the computational complexity of the Ising partition-function problem, or #Ising, and relate it to the logic-based counting of constraint-satisfaction problems, or #CSP. We show that known dichotomy results for #CSP give an easy proof of the hardness of #Ising and provide new intuition on where the difficulty of #Ising comes from. On the other hand, we also show that #Ising can be reduced to Weighted Model Counting (WMC). This enables us to take off-the-shelf model counters and apply them to #Ising. We show that this WMC approach outperforms state-of-the-art specialized tools for #Ising, thereby expanding the range of solvable problems in computational physics.

An association of the members, by the members, for the members

Abstract: C ON TRA R Y TO WHAT many of my non-U.S. colleagues believe, the “A” in ACM stands not for “American” but for “Association.” An association is a group of individuals who come together for a common purpose. An incorporated association is a legal entity that can do things in its own name, such as enter into contracts. So, ACM is a legal entity representing all ACM members. In other words, ACM is us! But an association of about 100,000 members cannot be run directly by its members. ACM’s Constitution and Bylaws specify its governance structure: Council, Executive Committee, President, and the like. Furthermore, an association with annual expenses of close to $50M requires the affairs of the association to be run by professional staff. The ACM Bylaws state that “There shall be an Executive Director and Chief Executive Officer who shall be a paid employee of the Association, ... and shall be responsible for the general administration of the affairs of the Association.” In May 2021, I wrote about “The Agency Trilemma and ACM,” referring to the gap between the members, who constitute the association, the elected officers, and the staff. I argued this gap may be partly responsible for the fact that the younger generation of computing professionals does not seem to feel the need to join ACM. While the tech industry, together with computing-degree-program enrollments, have been booming over the past decade, ACM membership growth has been modest at best. Accompanying the stagnant membership is, I suspect, the aging of the membership—a threat to the long-

DPSampler: Exact Weighted Sampling Using Dynamic Programming

Abstract: The problem of exact weighted sampling of solutions of Boolean formulas has applications in Bayesian inference, testing, and verification. The state-of-the-art approach to sampling involves carefully decomposing the input formula and compiling a data structure called d-DNNF in the process. Recent work in the closely connected field of model counting, however, has shown that smartly composing different subformulas using dynamic programming and Algebraic Decision Diagrams (ADDs) can outperform d-DNNF-style approaches on many benchmarks. In this work, we present a modular algorithm called DPSampler that extends such dynamic-programming techniques to the problem of exact weighted sampling. DPSampler operates in three phases. First, an execution plan in the form of a project-join tree is computed using tree decompositions. Second, the plan is used to compile the input formula into a succinct tree-of-ADDs representation. Third, this tree is traversed to generate a random sample. This decoupling of planning, compilation and sampling phases enables usage of specialized libraries for each purpose in a black-box fashion. Further, our novel ADD-sampling algorithm avoids the need for expensive dynamic memory allocation required in previous work. Extensive experiments over diverse sets of benchmarks show DPSampler is more scalable and versatile than existing approaches.

Divide-and-Conquer Determinization of Büchi Automata based on SCC Decomposition

Abstract: . The determinization of a nondeterministic B¨uchi automaton (NBA) is a fundamental construction of automata theory, with ap-plications to probabilistic verification and reactive synthesis. The stan-dard determinization constructions, such as the ones based on the Safra-Piterman’s approach, work on the whole NBA. In this work we propose a divide-and-conquer determinization approach. To this end, we first classify the strongly connected components (SCCs) of the given NBA as inherently weak, deterministic accepting, and nondeterministic accepting. We then present how to determinize each type of SCC independently from the others; this results in an easier handling of the determinization algorithm that takes advantage of the structure of that SCC. Once all SCCs have been determinized, we show how to compose them so to obtain the final equivalent deterministic Emerson-Lei automaton, which can be converted into a deterministic Rabin automaton without blow-up of states and transitions. We implement our algorithm in a our tool COLA and empirically evaluate COLA with the state-of-the-art tools Spot and Owl on a large set of benchmarks from the literature. The experimental results show that our prototype COLA outperforms Spot and Owl regarding the number of states and transitions.

Technology and democracy

Abstract: column Free AccessTechnology and democracy Author: Moshe Y. Vardi Rice University, Houston, TX Rice University, Houston, TXView Profile Authors Info & Claims Communications of the ACMVolume 65Issue 9September 2022 pp 5https://doi.org/10.1145/3550062Published:19 August 2022Publication History 0citation7,178DownloadsMetricsTotal Citations0Total Downloads7,178Last 12 Months7,178Last 6 weeks285 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteView all FormatsPDF

Satisfiability checking for Mission-time LTL (MLTL)

Abstract: Mission-time Linear Temporal Logic (LTL), abbreviated as MLTL, is a bounded variant of Metric Temporal Logic (MTL) over naturals designed to generically specify requirements for mission-based system operation common to aircraft, spacecraft, vehicles, and robots. Despite the utility of MLTL as a specification logic, major gaps remain in analyzing MLTL, e.g., for specification debugging or model checking, centering on the absence of any complete MLTL satisfiability checker. In this paper, we explore both the theoretical and algorithmic problems of MLTL satisfiability checking. We prove that the MLTL satisfiability checking problem is NEXPTIME-complete and that satisfiability checking MLTL0, the variant of MLTL where all intervals start at 0, is PSPACE-complete. To explore the best algorithmic solution for MLTL satisifiability checking, we reduce this problem to LTL satisfiability checking, LTLf (LTL over finite traces) satisfiability checking, and model checking respectively, thus conducting translations for MLTL-to-LTL, MLTL-to-LTLf, and MLTL-to-SMV. Moreover, we propose a new SMT-based solution for MLTL satisfiability checking and create a translation for MLTL-to-SMT. Our extensive experimental evaluation shows that while the MLTL-to-SMV translation with NuXmv model checker performs best on the benchmarks whose interval ranges are small (than 100), the MLTL-to-SMT translation with the Z3 SMT solver offers the most scalable performance.

Public and Private Affairs in Strategic Reasoning

Abstract: Do agents know each others’ strategies? In multi-process software construction, each process has access to the processes already constructed; but in typical human-robot interactions, a human may not announce its strategy to the robot (indeed, the human may not even know their own strategy). This question has often been overlooked when modeling and reasoning about multi-agent systems. In this work, we study how it impacts strategic reasoning. To do so we consider Strategy Logic (SL), a well-established and highly expressive logic for strategic reasoning. Its usual semantics, which we call “white-box semantics”, models systems in which agents “broadcast” their strategies. By adding imperfect information to the evaluation games for the usual semantics, we obtain a new semantics called “black-box semantics”, in which agents keep their strategies private. We consider the model-checking problem and show that the black-box semantics has much lower complexity than white-box semantics for an important fragment of Strategy Logic.