Fast learning of relational kernels

We develop a general theoretical framework for statistical logical learning with kernels based on dynamic propositionalization, where structure learning corresponds to inferring a suitable kernel on logical objects, and parameter learning corresponds to function learning in the resulting reproducing kernel Hilbert space. In particular, we study the case where structure learning is performed by a simple FOIL-like algorithm, and propose alternative scoring functions for guiding the search process. We present an empirical evaluation on several data sets in the single-task as well as in the multi-task setting.     

Statistical relational databases: normal forms

Problems associated with defining normal forms of relational tables relevant to statistical processing are discussed. The concepts of derived identifier, class identifier, derived class-counts, count domains, compact domains, and uniform domains for statistical relational tables are introduced. The structures of the first and the second statistical-normal forms and the relational decompositions needed to achieve them are also discussed. It is shown that the statistical-normal form can be an important method to determine whether the usual statistical analysis techniques are valid. Some suggestions are presented for extending the structured query language (SQL) statements to achieve these operations on statistical relational tables. Some results linking Codd's normal forms with statistical normal forms are discussed. Relational statistical abnormalities, called outlyers, are also discussed     

动态模糊逻辑关系学习

提出一种动态模糊逻辑（DFL）关系学习方法,该方法处理了动态模糊谓词和学习不同种类的动态模糊一阶规则的程序。针对不同类型的规则,定义了相关的置信度来考虑算法中的动态模糊谓词。通过实例验证了算法的有效性。     

Parameter learning of personalized trust models in broker-based distributed trust management

Distributed trust management addresses the challenges of eliciting, evaluating and propagating trust for service providers on the distributed network. By delegating trust management to brokers, individual users can share their feedbacks for services without the overhead of maintaining their own ratings. This research proposes a two-tier trust hierarchy, in which a user relies on her broker to provide reputation rating about any service provider, while brokers leverage their connected partners in aggregating the reputation of unfamiliar service providers. Each broker collects feedbacks from its users on past transactions. To accommodate individual differences, personalized trust is modeled with a Bayesian network. Training strategies such as the expectation maximization (EM) algorithm can be deployed to estimate both server reputation and user bias. This paper presents the design and implementation of a distributed trust simulator, which supports experiments under different configurations. In addition, we have conducted experiments to show the following. 1) Personal rating error converges to below 5% consistently within 10,000 transactions regardless of the training strategy or bias distribution. 2) The choice of trust model has a significant impact on the performance of reputation prediction. 3) The two-tier trust framework scales well to distributed environments. In summary, parameter learning of trust models in the broker-based framework enables both aggregation of feedbacks and personalized reputation prediction.

A taxonomy of weight learning methods for statistical relational learning

Machine Learning - Statistical relational learning (SRL) frameworks are effective at defining probabilistic models over complex relational data. They often use weighted first-order logical rules...     

Interactive relational reinforcement learning of concept semantics

We present a framework for the machine learning of denotational concept semantics using a simple form of symbolic interaction of machines with human users. The capability of software agents and robots to learn how to communicate verbally with human users would obviously be highly useful in several real-world applications, and our framework is meant to provide a further step towards this goal. Whereas the large majority of existing approaches to the machine learning of word sense and other language aspects focuses on learning using text corpora, our framework allows for the interactive learning of concepts in a dialog of human and agent, using an approach in the area of Relational Reinforcement Learning. Such an approach has a wide range of possible applications, e.g., the interactive acquisition of semantic categories for the Semantic Web, Human-Computer Interaction, (interactive) Information Retrieval, and Natural Language Processing.     

Gradient-based boosting for statistical relational learning: The relational dependency network case

Dependency networks approximate a joint probability distribution over multiple random variables as a product of conditional distributions. Relational Dependency Networks (RDNs) are graphical models that extend dependency networks to relational domains. This higher expressivity, however, comes at the expense of a more complex model-selection problem: an unbounded number of relational abstraction levels might need to be explored. Whereas current learning approaches for RDNs learn a single probability tree per random variable, we propose to turn the problem into a series of relational function-approximation problems using gradient-based boosting. In doing so, one can easily induce highly complex features over several iterations and in turn estimate quickly a very expressive model. Our experimental results in several different data sets show that this boosting method results in efficient learning of RDNs when compared to state-of-the-art statistical relational learning approaches.     

The role of trust in e-commerce relational exchange: A unified model

Recently, studies of B2C e-commerce have used intention theory to understand the role of trust of Internet transactions but most have investigated only a component of e-commerce (e.g., initial adoption or continuance) and neglected the role of good relations with the consumer in ensuring a successful sustained relationship. Therefore, a model that went beyond intention and included key relational concepts (satisfaction, value, loyalty, etc.) was developed. Trust and its components are a major part of this model, which was based on strong theoretical foundations. Fifteen hypotheses were formulated. Data on the constructs were collected from 420 respondents and analyzed using elliptical re-weighted least squares as the estimation method to test model validity and the hypotheses. An additional relationship between satisfaction and customer loyalty was investigated. Implications for researchers and practitioners are discussed.     

Translating networked learning: un-tying relational ties

Abstract   This article explores the social network of learning beyond a functional understanding of social relations. It describes and interprets the realities of networked learning within a particular postgraduate course in an English university setting. It draws attention to some of the limitations of the increasing interest in the use of social network analysis (SNA) alongside content analysis of recent studies in the field of networked learning. In particular, SNA has been used to analyse response relations among participants in online discussions in terms of, for example, density and centrality. It argues for a different approach to a network of learning, focusing on the relational effects of multiple technical and social arrangements and engagements beyond the response relations the online environment is able to capture and store. This approach emphasizes network processes rather than network structures.     

Distribution-based aggregation for relational learning with identifier attributes

Identifier attributes—very high-dimensional categorical attributes such as particular product ids or people's names—rarely are incorporated in statistical modeling. However, they can play an important role in relational modeling: it may be informative to have communicated with a particular set of people or to have purchased a particular set of products. A key limitation of existing relational modeling techniques is how they aggregate bags (multisets) of values from related entities. The aggregations used by existing methods are simple summaries of the distributions of features of related entities: e.g., MEAN, MODE, SUM, or COUNT. This paper's main contribution is the introduction of aggregation operators that capture more information about the value distributions, by storing meta-data about value distributions and referencing this meta-data when aggregating—for example by computing class-conditional distributional distances. Such aggregations are particularly important for aggregating values from high-dimensional categorical attributes, for which the simple aggregates provide little information. In the first half of the paper we provide general guidelines for designing aggregation operators, introduce the new aggregators in the context of the relational learning system ACORA (Automated Construction of Relational Attributes), and provide theoretical justification. We also conjecture special properties of identifier attributes, e.g., they proxy for unobserved attributes and for information deeper in the relationship network. In the second half of the paper we provide extensive empirical evidence that the distribution-based aggregators indeed do facilitate modeling with high-dimensional categorical attributes, and in support of the aforementioned conjectures. Editors: Hendrik Blockeel, David Jensen and Stefan Kramer An erratum to this article is available at .     

从关系数据库学习OWL本体的方法

提出了一种从关系数据库半自动学习OWL本体的方法.该方法在形式化表示关系数据库模式和OWL本体的基础上,遵循从关系数据库模式到OWL本体的一组通用映射方法和规则,并基于Java 2平台实现了原型工具OntoLeamer.利用OntoLeamer进行的典型案例研究表明了该方法的有效性.     

Applying the information bottleneck to statistical relational learning

In this paper we propose to apply the Information Bottleneck (IB) approach to the sub-class of Statistical Relational Learning (SRL) languages that are reducible to Bayesian networks. When the resulting networks involve hidden variables, learning these languages requires the use of techniques for learning from incomplete data such as the Expectation Maximization (EM) algorithm. Recently, the IB approach was shown to be able to avoid some of the local maxima in which EM can get trapped when learning with hidden variables. Here we present the algorithm Relational Information Bottleneck (RIB) that learns the parameters of SRL languages reducible to Bayesian Networks. In particular, we present the specialization of RIB to a language belonging to the family of languages based on the distribution semantics, Logic Programs with Annotated Disjunction (LPADs). This language is prototypical for such a family and its equivalent Bayesian networks contain hidden variables. RIB is evaluated on the IMDB, Cora and artificial datasets and compared with LeProbLog, EM, Alchemy and PRISM. The experimental results show that RIB has good performances especially when some logical atoms are unobserved. Moreover, it is particularly suitable when learning from interpretations that share the same Herbrand base.     

A method for reduction of examples in relational learning

Feature selection methods often improve the performance of attribute-value learning. We explore whether also in relational learning, examples in the form of clauses can be reduced in size to speed up learning without affecting the learned hypothesis. To this end, we introduce the notion of safe reduction: a safely reduced example cannot be distinguished from the original example under the given hypothesis language bias. Next, we consider the particular, rather permissive bias of bounded treewidth clauses. We show that under this hypothesis bias, examples of arbitrary treewidth can be reduced efficiently. We evaluate our approach on four data sets with the popular system Aleph and the state-of-the-art relational learner nFOIL. On all four data sets we make learning faster in the case of nFOIL, achieving an order-of-magnitude speed up on one of the data sets, and more accurate in the case of Aleph.     

Statistical guidance in symbolic learning

Concept learning methods of Artificial Intelligence (AI) are usefully guided by statistical measures of concept quality. We review the application of statistical measures intutored methods oflearning from examples, describe the recent application of these measures toconceptual clustering, and propose statistical applications inexplanation-based learning.This work was supported by a Vanderbilt Research Council Grant.     

Statistical learning based facial animation

To synthesize real-time and realistic facial animation, we present an effective algorithm which combines image- and geometry-based methods for facial animation simulation. Considering the numerous motion units in the expression coding system, we present a novel simplified motion unit based on the basic facial expression, and construct the corresponding basic action for a head model. As image features are difficult to obtain using the performance driven method, we develop an automatic image feature recognition method based on statistical learning, and an expression image semi-automatic labeling method with rotation invariant face detection, which can improve the accuracy and efficiency of expression feature identification and training. After facial animation redirection, each basic action weight needs to be computed and mapped automatically. We apply the blend shape method to construct and train the corresponding expression database according to each basic action, and adopt the least squares method to compute the corresponding control parameters for facial animation. Moreover, there is a pre-integration of diffuse light distribution and specular light distribution based on the physical method, to improve the plausibility and efficiency of facial rendering. Our work provides a simplification of the facial motion unit, an optimization of the statistical training process and recognition process for facial animation, solves the expression parameters, and simulates the subsurface scattering effect in real time. Experimental results indicate that our method is effective and efficient, and suitable for computer animation and interactive applications.     

Statistical profiles of highly-rated learning objects

The continuously growth of learning resources available in on-line repositories has raised the concern for the development of automated methods for quality assessment. The current existence of on-line evaluations in such repositories has opened the possibility of searching for statistical profiles of highly-rated resources that can be used as priori indicators of quality. In this paper, we analyzed 35 metrics in learning objects refereed inside the MERLOT repository and elaborated profiles for these resources regarding the different categories of disciplines and material types available. We found that some of the intrinsic metrics presented significant differences between highly rated and poorly-rated resources and that those differences are dependent on the category of discipline to which the resource belongs and on the type of the resource. Moreover, we found that different profiles should be identified according to the type of rating (peer-review or user) under evaluation. At last, we developed an initial model using linear discriminant analysis to evaluate the strength of relevant metrics when performing an automated quality classification task. The initial results of this work are promising and will be used as the foundations for the further development of an automated tool for contextualized quality assessment of learning objects inside repositories.     

Graph kernels and Gaussian processes for relational reinforcement learning

RRL is a relational reinforcement learning system based on Q-learning in relational state-action spaces. It aims to enable agents to learn how to act in an environment that has no natural representation as a tuple of constants. For relational reinforcement learning, the learning algorithm used to approximate the mapping between state-action pairs and their so called Q(uality)-value has to be very reliable, and it has to be able to handle the relational representation of state-action pairs. In this paper we investigate the use of Gaussian processes to approximate the Q-values of state-action pairs. In order to employ Gaussian processes in a relational setting we propose graph kernels as a covariance function between state-action pairs. The standard prediction mechanism for Gaussian processes requires a matrix inversion which can become unstable when the kernel matrix has low rank. These instabilities can be avoided by employing QR-factorization. This leads to better and more stable performance of the algorithm and a more efficient incremental update mechanism. Experiments conducted in the blocks world and with the Tetris game show that Gaussian processes with graph kernels can compete with, and often improve on, regression trees and instance based regression as a generalization algorithm for RRL. Editors: David Page and Akihiro Yamamoto