The digitalization of music and the instant availability of millions of tracks on the Internet require new approaches to support the user in the exploration of these huge music collections. One possible approach to address this problem, which can also be found on popular online music platforms, is the use of user-created or automatically generated playlists (mixes). The automated generation of such playlists represents a particular type of the music recommendation problem with two special characteristics. First, the tracks of the list are usually consumed immediately at recommendation time; secondly, songs are listened to mostly in consecutive order so that the sequence of the recommended tracks can be relevant. In the past years, a number of different approaches for playlist generation have been proposed in the literature. In this paper, we review the existing core approaches to playlist generation, discuss aspects of appropriate offline evaluation designs and report the results of a comparative evaluation based on different datasets. Based on the insights from these experiments, we propose a comparably simple and computationally tractable new baseline algorithm for future comparisons, which is based on track popularity and artist information and is competitive with more sophisticated techniques in our evaluation settings.

When solving extensive-form games with large action spaces, typically significant abstraction is needed to make the problem manageable from a modeling or computational perspective. When this occurs, a procedure is needed to interpret actions of the opponent that fall outside of our abstraction (by mapping them to actions in our abstraction). This is called an action translation mapping. Prior action translation mappings have been based on heuristics without theoretical justification. We show that the prior mappings are highly exploitable and that most of them violate certain natural desiderata. We present a new mapping that satisfies these desiderata and has significantly lower exploitability than the prior mappings. Furthermore, we observe that the cost of this worst-case performance benefit (low exploitability) is not high in practice; our mapping performs competitively with the prior mappings against no-limit Texas Hold'em agents submitted to the 2012 Annual Computer Poker Competition. We also observe several paradoxes that can arise when performing action abstraction and translation; for example, we show that it is possible to improve performance by including suboptimal actions in our abstraction and excluding optimal actions.

An important aspect of agent evaluation in stochastic games, especially poker, is the need to reduce the outcome variance in order to get accurate and significant results. The current method used in the Annual Computer Poker Competition’s analysis is that of duplicate poker, an approach that leverages the ability to deal sets of cards to agents in order to reduce variance. This work explores a different approach to variance reduction by using a control variate based approach known as baseline. The baseline approach involves using an agent’s outcome in self play to create an unbiased estimator for use in agent evaluation and has been shown to work well in both poker and trading agent competition domains. Base- line does not require that the agents are able to be dealt sets of cards, making it a more robust technique than duplicate. This approach is compared to the current duplicate method, as well as other variations of duplicate poker on the results of the 2011 two player no-limit and three player limit Texas Hold’em ACPC tournaments.

Non-negative tensor factorization (NTF) has attracted great attention in the machine learning community. In this paper, we extend traditional non-negative tensor factorization into a supervised discriminative decomposition, referred as Supervised Non-negative Tensor Factorization with Maximum-Margin Constraint(SNTFM2). SNTFM2 formulates the optimal discriminative factorization of non-negative tensorial data as a coupled least-squares optimization problem via a maximum-margin method. As a result, SNTFM2 not only faithfully approximates the tensorial data by additive combinations of the basis, but also obtains a strong generalization power to discriminative analysis (in particularfor classification in this paper). The experimental results show the superiority of our proposed model over state-of-the-art techniques on both toy and real world data sets.

In many tasks related to reasoning about consequences of a logical theory, it is desirable to decompose the theory into a number of components with weakly-related or independent signatures. This facilitates reasoning when signature of a query formula belongs to only one of the components. However, an initial theory may be subject to change due to execution of actions affecting features mentioned in the theory. Having once computed a decomposition of a theory, one would like to know whether a decomposition has to be computed again for the theory obtained from taking into account the changes resulting from execution of an action. In the paper, we address this problem in the scope of the situation calculus, where change of an initial theory is related to the well-studied notion of progression. Progression provides a form of forward reasoning; it relies on forgetting values of those features which are subject to change and computing new values for them. We prove new results about properties of decomposition components under forgetting and show when a decomposition can be preserved in progression of an initial theory.

This paper proposes an effective approach to model the emotional space of words to infer their Sense Sentiment Similarity (SSS). SSS reflects the distance between the words regarding their senses and underlying sentiments. We propose a probabilistic approach that is built on a hidden emotional model in which the basic human emotions are considered as hidden. This leads to predict a vector of emotions for each sense of the words, and then to infer the sense sentiment similarity. The effectiveness of the proposed approach is investigated in two Natural Language Processing tasks: Indirect yes/no Question Answer Pairs Inference and Sentiment Orientation Prediction.

The tractability of a Constraint Satisfaction Problem (CSP)is guaranteed by a direct relationship between its consistencylevel and a structural parameter of its constraint network suchas the treewidth. This result is not widely exploited in practicebecause enforcing higher-level consistencies can be costlyand can change the structure of the constraint network andincrease its width. Recently, R(*,m)C was proposed as a relational consistency property that does not modify the structureof the graph and, thus, does not affect its width. In this paper,we explore two main strategies, based on a tree decomposition of the CSP, for improving the performance of enforcingR(*,m)C and getting closer to the above tractability condition. Those strategies are: a) localizing the application ofthe consistency algorithm to the clusters of the tree decomposition, and b) bolstering constraint propagation betweenclusters by adding redundant constraints at their separators,for which we propose three new schemes. We characterizethe resulting consistency properties by comparing them, theoretically and empirically, to the original R(*,m)C and thepopular GAC and maxRPWC, and establish the benefits ofour approach for solving difficult problems.

Given a classification task, what is the best way to teach the resulting boundary to a human? While machine learning techniques can provide excellent methods for finding the boundary, including the selection of examples in an online setting, they tell us little about how we would teach a human the same task. We propose to investigate the problem of example selection and presentation in the context of teaching humans, and explore a variety of mechanisms in the interests of finding what may work best. In particular, we begin with the baseline of random presentation and then examine combinations of several mechanisms: the indication of an example’s relative difficulty, the use of the shaping heuristic from the cognitive science literature (moving from easier examples to harder ones), and a novel kernel-based “coverage model” of the subject’s mastery of the task. From our experiments on 54 human subjects learning and performing a pair of synthetic classification tasks via our teaching system, we found that we can achieve the greatest gains with a combination of shaping and the coverage model.

Adaptive strategies are popular in poker AI research. Three desirable properties of an adaptive poker bot are optimality, generality and speed of response to new opponents. These three properties though cannot be achieved simultaneously for most imperfect information games; some trade-off must be made between them. This general principle is connected to recent work on poker AI and particularly the total bankroll competitions at the Annual Computer Poker Competition (ACPC). This paper is meant to generate discussion between researchers about how to explain and quantify these trade-offs better and to possible future directions for research.

Despite committees and elections being widespread in thereal-world, the design of agents for operating in humancomputer committees has received far less attention than thetheoretical analysis of voting strategies. We address this gapby providing an agent design that outperforms other voters ingroups comprising both people and computer agents. In oursetting participants vote by simultaneously submitting a ranking over a set of candidates and the election system uses a social welfare rule to select a ranking that minimizes disagreements with participants’ votes. We ran an extensive studyin which hundreds of people participated in repeated votingrounds with other people as well as computer agents that differed in how they employ strategic reasoning in their votingbehavior. Our results show that over time, people learn todeviate from truthful voting strategies, and use heuristics toguide their play, such as repeating their vote from the previous round. We show that a computer agent using a bestresponse voting strategy was able to outperform people in thegame. Our study has implication for agent designers, highlighting the types of strategies that enable agents to succeedin committees comprising both human and computer participants. This is the first work to study the role of computeragents in voting settings involving both human and agent participants.