Partially observable stochastic games (POSGs) are an attractive model for many multi-agent domains, but are computationally extremely difficult to solve. We present a new model, Markov games of incomplete information (MGII) which imposes a mild restriction on POSGs while overcoming their primary computational bottleneck. Finally we show how to convert a MGII into a continuous but bounded fully observable stochastic game. MGIIs represents the most general tractable model for multi-agent reinforcement learning to date.

Frequently, it is advantageous for an agent to model other agents in order to predict their behavior during an interaction. Modeling others as rational has a long tradition in AI and game theory, but modeling other agents’ departures from rationality is difficult and controversial. This paper proposes that bounded rationality be modeled as errors the agent being modeled is making while deciding on its action. We are motivated by the work on quantal response equilibria in behavioral game theory which uses Nash equilibria as the solution concept. In contrast, we use decision-theoretic maximization of expected utility. Quantal response assumes that a decision maker is rational, i.e., is maximizing his expected utility, but only approximately so, with an error rate characterized by a single error parameter. Another agent’s error rate may be unknown and needs to be estimated during an interaction. We show that the error rate of the quantal response can be estimated using Bayesian update of a suitable conjugate prior, and that it has a finitely dimensional sufficient statistic under strong simplifying assumptions. However, if the simplifying assumptions are relaxed, the quantal response does not admit a finite sufficient statistic and a more complex update is needed. This confirms the difficulty of using simple models of bounded rationality in general settings.

Having discovered a decision principle for a well-known prima facie duty theory in biomedical ethics to resolve particular cases of a common type of ethical dilemma, we developed three applications: a medical ethics advisor system, a medication reminder system and an instantiation of this system in a Nao robot. We are now developing a general, automated method for generating from scratch the ethics needed for a machine to function in a particular domain, without making the assumptions used in our prototype systems.

Existing crowdsourcing markets are often inaccessible to workers living at the bottom of the economic pyramid. We present MobileWorks, a mobile phone-based crowdsourcing platform intended to provide employment to developing world users. MobileWorks provides human optical character recognition (OCR) tasks that can be completed by workers on low-end mobile phones through a web browser. To address the limited screen resolution available on low-end phones, MobileWorks divides documents into many small pieces and sends each piece to a different worker. An initial pilot study with 10 users over a two month period revealed that it is feasible to do basic OCR tasks using a simple mobile web-based application. We find that workers using MobileWorks average 120 tasks per hour at an accuracy rate of 99% using a multiple entry solution. In addition, users had a positive experience with MobileWorks: all study participants would recommend MobileWorks to friends and family.

Mutual exclusions provide useful information for learn- ing classes of concepts. We designed MuSweeper as a MineSweeper-like game to collect mutual exclusions from web users. Using the mechanism of an exten- sive game with Imperfect information, our experiments showed MuSweeper to collect mutual exclusions with high precision and efficiency.

Crowdsourcing platforms, such as Amazon Mechanical Turk, have enabled the construction of scalable applications for tasks ranging from product categorization and photo tagging to audio transcription and translation. These vertical applications are typically realized with complex, self-managing workflows that guarantee quality results. But constructing such workflows is challenging, with a huge number of alternative decisions for the designer to consider. We argue the thesis that “Artificial intelligence methods can greatly simplify the process of creating and managing complex crowdsourced workflows.” We present the design of CLOWDER, which uses machine learning to continually refine models of worker performance and task difficulty. Using these models, CLOWDER uses decision-theoretic optimization to 1) choose between alternative workflows, 2) optimize parameters for a workflow, 3) create personalized interfaces for individual workers, and 4) dynamically control the workflow. Preliminary experience suggests that these optimized workflows are significantly more economical (and return higher quality output) than those generated by humans.

We describe a framework for rapidly prototyping applications which require intelligent visual processing, but for which reliable algorithms do not yet exist, or for which engineering those algorithms is too costly. The framework, CrowdSight, leverages the power of crowdsourcing to offload intelligent processing to humans, and enables new applications to be built quickly and cheaply, affording system builders the opportunity to validate a concept before committing significant time or capital. Our service accepts requests from users either via email or simple mobile applications, and handles all the communication with a backend human computation platform. We build redundant requests and data aggregation into the system freeing the user from managing these requirements. We validate our framework by building several test applications and verifying that prototypes can be built more easily and quickly than would be the case without the framework.

We present techniques for gathering data that expose errors of automatic predictive models. In certain common settings, traditional methods for evaluating predictive models tend to miss rare-but-important errors---most importantly, rare cases for which the model is confident of its prediction (but wrong). In this paper we present a system that, in a game-like setting, asks humans to identify cases what will cause the predictive-model-based system to fail. Such techniques are valuable in discovering problematic cases that do not reveal themselves during the normal operation of the system, and may include cases that are rare but catastrophic. We describe the design of the system, including design iterations that did not quite work. In particular, the system incentivizes humans to provide examples that are difficult for the model to handle, by providing a reward proportional to the magnitude of the predictive model's error. The humans are asked to ``\emph{Beat the Machine}'' and find cases where the automatic model (``\emph{the Machine}'') is wrong. Experiments show that the humans using Beat the Machine identify more errors than traditional techniques for discovering errors in from predictive models, and indeed, they identify many more errors where the machine is confident it is correct. Further, the cases the humans identify seem to be not simply outliers, butcoherent areas missed completely by the model. Beat the machine identifies the ``unknown unknowns.''

The strategies seem designed to mitigate guilt over the parents' role in their The GenIE Assistant is an implemented proof-of-concept child's inheritance of a genetic condition. The names used computational model of normative biomedical argument to refer to the strategies in this paper and examples of generation informed by study of a corpus of letters each are listed below. All four apply to cases of written by genetic counselors to their clients (Green et al. autosomal recessive inheritance, while only the first two 2011). The goal of the model is to generate transparent apply to cases of autosomal dominant inheritance.

This paper describes the development of a corpus of blog posts that are annotated for the presence of attempts to persuade and corresponding tactics employed in persuasive messages. We investigate the feasibility of classifying blog posts as persuasive or non-persuasive on the basis of lexical features in the text and the tactics (as provided by human annotators). Annotated tactics provide substantial assistance in classifying persuasion, particularly tactics indicating formal reasoning, deontic obligation, and discussions of possible outcomes, suggesting that learning to identify tactics may be an excellent first step to detecting attempts to persuade.