The utility of including cyclic flows of control in plans has been long recognized by the planning community. Loops in a plan help increase both its applicability and the compactness of representation. However, progress in finding such plans has been limited largely due to lack of methods for reasoning about the correctness and safety properties of loops of actions. We present an overview of recent results for determining the class of problems that a plan with loops can solve. These methods can be used to direct the construction of a rich new form of generalized plans that solve a desired class of problems.

Information about the spread of crop disease is vital in developing countries, and as a result the governments of such countries devote scarce resources to gathering such data. Unfortunately, current surveys tend to be slow and expensive, and hence also tend to gather insufficient quantities of data. In this work we describe three general methods for improving the use of survey resources by performing data collection with mobile devices and by directing survey progress through the application of AI techniques. First, we describe a spatial disease density model based on Gaussian process ordinal regression, which offers a better representation of the disease level distribution, as compared to the statistical approaches typically applied. Second, we show how this model can be used to dynamically route survey teams to obtain the most valuable survey possible given a fixed budget. Third, we demonstrate that the diagnosis of plant disease can be automated using images taken by a camera phone, enabling data collection by survey workers with only basic training. We have applied our methods to the specific challenge of viral cassava disease monitoring in Uganda, for which we have implemented a real-time mobile survey system that will soon see practical use.

Many people read online reviews written by other users to learn more about a product or venue. However, the overwhelming amount of user- generated reviews and variance in length, detail and quality across the reviews make it difficult to glean useful information. In this paper, we present a summarization system called Review Spotlight. It provides a brief overview of reviews by using adjective- noun word pairs extracted from the review text. The system also allows the user to click any word pair to read the original sentences from which the word pair was extracted. We present our system implementation as a Google Chrome browser extension, and an evaluation on how two word pair scoring methods (TF and TF-IDF) affect the identification of useful word pairs.

Opinion leaders play an important role in influencing people’s beliefs, actions and behaviors. Although a number of methods have been proposed for identifying influentials using secondary sources of information, the use of primary sources, such as surveys, is still favored in many domains. In this work we present a new surveying method which combines secondary data with partial knowledge from primary sources to guide the information gathering process. We apply our proposed active surveying method to the problem of identifying key opinion leaders in the medical field, and show how we are able to accurately identify the opinion leaders while minimizing the amount of primary data required, which results in significant cost reduction in data acquisition without sacrificing its integrity.

In the past few years, sentiment analysis and opinion mining becomes a popular and important task. These studies all assume that their opinion resources are real and trustful. However, they may encounter the faked opinion or opinion spam problem. In this paper, we study this issue in the context of our product review mining system. On product review site, people may write faked reviews, called review spam, to promote their products, or defame their competitors' products. It is important to identify and filter out the review spam. Previous work only focuses on some heuristic rules, such as helpfulness voting, or rating deviation, which limits the performance of this task. In this paper, we exploit machine learning methods to identify review spam. Toward the end, we manually build a spam collection from our crawled reviews. We first analyze the effect of various features in spam identification. We also observe that the review spammer consistently writes spam. This provides us another view to identify review spam: we can identify if the author of the review is spammer. Based on this observation, we provide a two-view semi-supervised method, co-training, to exploit the large amount of unlabeled data. The experiment results show that our proposed method is effective. Our designed machine learning methods achieve significant improvements in comparison to the heuristic baselines.

This paper investigates belief revision where the underlying logic is that governing Horn clauses. It proves to be the case that classical (AGM) belief revision doesn’t immediately generalise to the Horn case. In particular, a standard construction based on a total preorder over possible worlds may violate the accepted (AGM) postulates. Conversely, Horn revision functions in the obvious extension to the AGM approach are not captured by total preorders over possible worlds. We address these difficulties by first restricting the semantic construction to "well behaved" orderings; and second, by augmenting the revision postulates by an additional postulate. This additional postulate is redundant in the AGM approach but not in the Horn case. In a representation result we show that these two approaches coincide. Arguably this work is interesting for several reasons. It extends AGM revision to inferentially-weaker Horn theories; hence it sheds light on the theoretical underpinnings of belief change, as well as generalising the AGM paradigm. Thus, this work is relevant to revision in areas that employ Horn clauses, such as deductive databases and logic programming, as well as areas in which inference is weaker than classical logic, such as in description logic.

Given the number of online sources for news, the volumes of news generated are so daunting that gaining insight from these collections become impossible without some aid to link them. Semantic linking of news articles facilitates grouping of similar or relevant news stories together for ease of human consumption. For example, a political analyst may like to have a single view of all news articles that report visits of State heads of different countries to a single country to make an in-depth analytical report on the possible impacts of all associated events. It is likely that no news source links all the relevant news together. In this paper, we discuss a multi-resolution, multi-perspective news analysis system that can link news articles collected from diverse sources over a period of time. The distinctive feature of the proposed news linking system is its capability to simultaneously link news articles and stories at multiple levels of granularity. At the lowest level several articles reporting the same event are linked together. Higher level groupings are more contextual and semantic. We have deployed a range of algorithms that use statistical text processing and Natural Language Processing techniques. The system is incremental in nature and depicts how stories have evolved over time along with main actors and activities. It also illustrates how a single story diverges into multiple themes or multiple stories converge due to conceptual similarity. Accuracy of linking thematically and conceptually linked news articles are also presented.

This paper identifies a widely existing phenomenon in web data, which we call the "words of few mouths" phenomenon. This phenomenon, in the context of online reviews, refers to the case that a large fraction of the reviews are each voted only by very few users. We discuss the challenges of "words of few mouths" in the development of recommender systems based on users' opinions and advocate probabilistic methodologies to handle such challenges. We develop a probabilistic model and correspondingly a logistic regression based learning algorithm for review helpfulness prediction. Our experimental results indicate that the proposed model outperforms the current state-of-the-art algorithms not only in the presence of the "words of few mouths" phenomenon, but also in the absence of such phenomena.

We describe an innovative application of a novel game-theoretic approach for a \textit{national scale} security deployment. Working with the United States Transportation Security Administration (TSA), we have developed a new application called GUARDS to allocate the TSA's limited resources across hundreds of security activities to provide protection at over 400 United States airports. Similar security applications (e.g., ARMOR and IRIS) have focused on one-off tailored applications and one security activity (e.g. checkpoints) per application, GUARDS on the other hand faces three new key issues: (i) reasoning about hundreds of heterogeneous security activities; (ii) reasoning over diverse potential threats; (iii) developing a system designed for hundreds of end-users. Since a national deployment precludes tailoring to specific airports, our key ideas are: (i) creating a new game-theoretic framework that allows for heterogeneous defender activities and compact modeling of a large number of threats; (ii) developing an efficient solution technique based on general purpose Stackelberg game solvers; (iii) taking a partially centralized approach for knowledge acquisition. The scheduling assistant has been delivered to the TSA and is currently undergoing evaluation for scheduling practices at an undisclosed airport. If successful, the TSA intends to incorporate the system into their unpredictable scheduling practices nationwide.

Humans have developed jurisprudence as a mechanism to solve conflictive situations by using past experiences. Following this principle, we propose an approach to enhance a multi-agent system by adding an authority which is able to generate new regulations whenever conflicts arise. Regulations are generated by learning from previous similar situations, using a machine learning technique (based on Case-Based Reasoning) that solves new problems using previous experiences. This approach requires: to be able to gather and evaluate experiences; and to be described in such a way that similar social situations require similar regulations. As a scenario to evaluate our proposal, we use a simplified version of a traffic scenario, where agents are traveling cars. Our goals are to avoid collisions between cars and to avoid heavy traffic. These situations, when happen, lead to the synthesis of new regulations. At each simulation step, applicable regulations are evaluated in terms of their effectiveness and necessity. Overtime the system generates a set of regulations that, if followed, improve system performance (i.e. goal achievement).