For understanding humorous dialogue, a collection of humorous expressions is needed. In addition to humorous expressions, their annotations are important to be used as language resources. In this paper, we analyzed how human assessors annotate humorous expressions extracted from an online community-based question-answering (CQA) corpus, which contains many interesting examples of humorous communication. We analyzed the annotation results of a collection of humorous expressions as done by 28 annotators in terms of the degree of humor and categorization of humor. We found the assessments to be quite subjective, and only marginal inter-annotator agreements were observed. This result suggests that the variability in humor annotations is not noise resulting from erroneous assessment but is rooted in personality differences of the annotators. It would be necessary to incorporate the individual differences in humor perception for properly utilizing the resources. We discuss the possibility to improve the collection process by applying filtering techniques.

We study the mixtures of factorizing probability distributions represented as visible marginal distributions in stochastic layered networks. We take the perspective of kernel transitions of distributions, which gives a unified picture of distributed representations arising from Deep Belief Networks (DBN) and other networks without lateral connections. We describe combinatorial and geometric properties of the set of kernels and products of kernels realizable by DBNs as the network parameters vary. We describe explicit classes of probability distributions, including exponential families, that can be learned by DBNs. We use these submodels to bound the maximal and the expected Kullback-Leibler approximation errors of DBNs from above depending on the number of hidden layers and units that they contain.

Although decision processes are often described at the individual level of cognition (e.g. Tversky and Kahnemann The particular beliefs instantiated within the model are (1981)), they are subject to social and cultural influences based on a combination of results from empirical studies at both the interpersonal and societal levels. The adoption of technology adoption by Venkatesh et al. (2003). of new technology depends on various factors, such The UTAUT model combines eight of the most prominent as the type of technology, the context or culture in which technology-acceptance models observed in the literature and the technology is introduced, and the individual decisions provides a definitive list of variables that are critically relevant by people within that culture, as most individuals evaluate to an individual's Behavioral Intention (BI) and Use Behavior an innovation from the subjective evaluations of peers who (UB) for adopting a new technology, including Performance have adopted an innovation (see Watts and Dodds (2007) Expectancy (PE), Effort Expectancy (EE), Social for a discussion of network-diffused influence). These influences Influence (SI), Facilitating Conditions (FC), and Voluntariness propagate through the social network as a function of Use (VoU). of agent interactions.

Many strategic actions carry a "contagious" component beyond the immediate locale of the effort itself. Viral marketing and peacekeeping operations have both been observed to have a spreading effect. In this work, we use counterinsurgency as our illustrative domain. Defined as the effort to block the spread of support for an insurgency, such operations lack the manpower to defend the entire population and must focus on the opinions of a subset of local leaders. As past researchers of security resource allocation have done, we propose using game theory to develop such policies and model the interconnected network of leaders as a graph. Unlike this past work in security games, actions in these domains possess a probabilistic, non-local impact. To address this new class of security games, recent research has used novel heuristic oracles in a double oracle formulation to generate mixed strategies. However, these heuristic oracles were evaluated only on runtime and quality scaling with the graphsize. Given the complexity of the problem, numerous other problem features and metrics must be considered to better inform practical application of such techniques. Thus, this work provides a thorough experimental analysis including variations of the contagion probability average and standard deviation. We extend the previous analysis to also examine the size of the action set constructed in the algorithms and the final mixed strategies themselves. Our results indicate that game instances featuring smaller graphs and low contagion probabilities converge slowly while games with larger graphs and medium contagion probabilities converge most quickly.

There has been a rapid proliferation of location-based social networks (LBSNs) during the last years. The spatial component of these systems provides a rich source of information that can be exploited by a number of novel services. However, to better design such services, it is important to understand the way people make use of these platforms and how this usage changes over time. While there exist studies that examine the motivations of people for adopting the usage of LBSNs and the temporal dynamics of these motivations, they are based on interviews and are mostly qualitative. Motivations can further only indirectly reveal or help us infer user behavior. In this paper, we analyze data from two commercial LBSNs to examine the temporal evolution of usage patterns to see what the data on their own reveal. We nd that users of two social networks that we examined increase their level of activity as they use the system. However, depending on the main purpose of the underlying LBSN, users may exhibit dierent behaviors over time. We believe that our ndings can open new directions and stimulate further research on areas such as location prediction and its applications (e.g., urban and transportation planning and location-based advertisment).

Network analysis is a growing field across many domains, including computer vision, social media marketing, transportation networks, and intelligence analysis. The growing use of digital communication devices and platforms, as well as persistent surveillance sensors, has resulted in explosion of the quantity of data and stretched the abilities of current technologies to process this data and draw meaningful conclusions. Current tools either require significant levels of manual intervention (e.g., to prepare the data, to define patterns, or to draw conclusions from data) or are unable to generalize to new data sources and analysis needs. In this paper, we present automated solutions to two major problems in network analysis: (a) finding patterns in the network data that contains high levels of noise and irrelevant information; and (b) learning repetitive patterns and dependencies between entities and attributes. Our modeling framework represents network data using multi-attributed graphs that can encode various discrete and continuous features and relationships between network entities. The pattern search and learning model is based on probabilistic multi-attributed graph matching, and implemented using distributed message passing algorithms. Our algorithms achieved high accuracy rates in learning and finding patterns in the data, are flexible to new domains and data types, and scale to large datasets using the Map-Reduce framework.

In the real world we observe a proliferation of regional dialects and jargons. Most of the research on naming conventions focuses on how to explain the process that allows a single naming convention to establish itself. This paper presents a different approach that aims to investigate why different conventions may emerge and coexist for a certain amount of time. The naming game is an abstraction of lexical acquisition dynamics, in which n agents try to find an agreement on the names to give to objects. To understand how different heterogeneous conventions emerge, I discuss a naming game model that takes into account experimental data on human and animal learning.

We consider the problem of extending manually trainedagents via evaluative reinforcement (TAMER) in con-tinuous state and action spaces. The early work TAMERframework allows a non-technical human to train anagent through a natural form of human feedback, neg-ative or positive. The advantages of TAMER havebeen shown on applications such as training Tetris andMountain Car with only human feedback, Cart-poleand Mountain Car with human feedback and environ-ment reward (augmenting reinforcement learning withhuman feedback). However, those methods are origi-nally designed for discrete state-action, or continuousstate-discrete action problems. In this paper, we intro-duce an extension of TAMER to allow both continu-ous states and actions. The new scheme, actor-criticTAMER, extends the original TAMER to allow usingany general function approximation of a human trainer’sreinforcement signal. Our extension still allows rein-forcement learning to be easily combined with humanfeedback. The experimental results show that the pro-posed method helps a human trainer successfully trainan agent in two continuous state-action domains: Moun-tain Car, and Cart-pole (balancing).

Most learning from demonstration algorithms are implemented with a certain set of variables that are known to be important for the agent. The agent is hardcoded to use those variables for learning the task (or a set of parameters). In this work we try to understand the causal structure of a demonstrated task in order to find: which variables cause what other variables to change, and which variables are independent from the others. We used a realistic simulator to record a simple pick and place task demonstration data, and recovered different causal models using the data in Tetrad, a computer program that searches for causal and statistical models. Our findings show that it is possible to deduce irrelevant variables to a demonstrated task, using the recovered causal structure.

Members of a multi-robot team, operating within close quarters, need to avoid crashing into each other. Simple collision avoidance methods can be used to prevent such collisions, typically by computing the distance to other robots and stopping, perhaps moving away, when this distance falls below a certain threshold. While this approach may avoid disaster, it may also reduce the team's efficiency if robots halt for a long time to let others pass by or if they travel further to move around one another. This paper reports on experiments where a human operator, through a graphical user interface, watches robots perform an exploration task. The operator can manually suspend robots' movements before they crash into each other, and then resume their movements when their paths are clear. Experiment logs record the robots' states when they are paused and resumed. A behavior pattern for collision avoidance is learned, by classifying the states of the robots' environment when the human operator issues "wait" and "resume" commands. Preliminary results indicate that it is possible to learn a classifier which models these behavior patterns, and that different human operators consider different factors when making decisions about stopping and starting robots.