In recent years, neural machine translation (NMT) has demonstrated state-of-the-art machine translation (MT) performance. It is a new approach to MT, which tries to learn a set of parameters to maximize the conditional probability of target sentences given source sentences. In this paper, we present a novel approach to improve the translation performance in NMT by conveying topic knowledge during translation. The proposed topic-informed NMT can increase the likelihood of selecting words from the same topic and domain for translation. Experimentally, we demonstrate that topic-informed NMT can achieve a 1.15 (3.3% relative) and 1.67 (5.4% relative) absolute improvement in BLEU score on the Chinese-to-English language pair using NIST 2004 and 2005 test sets, respectively, compared to NMT without topic information.

Neural network training has been shown to be advantageous in many natural language processing applications, such as language modelling or machine translation. In this paper, we describe in detail a novel domain adaptation mechanism in neural network training. Instead of learning and adapting the neural network on millions of training sentences – which can be very time-consuming or even infeasible in some cases – we design a domain adaptation gating mechanism which can be used in recurrent neural networks and quickly learn the out-of-domain knowledge directly from the word vector representations with little speed overhead. In our experiments, we use the recurrent neural network language model (LM) as a case study. We show that the neural LM perplexity can be reduced by 7.395 and 12.011 using the proposed domain adaptation mechanism on the Penn Treebank and News data, respectively. Furthermore, we show that using the domain-adapted neural LM to re-rank the statistical machine translation n-best list on the French-to-English language pair can significantly improve translation quality.

Several anonymization techniques, such as generalization and bucketization, have been designed for privacy preserving microdata publishing. Recent work has shown that generalization loses considerable amount of information, especially for high-dimensional data. Bucketization, on the other hand, does not prevent membership disclosure and does not apply for data that do not have a clear separation between quasi-identifying attributes and sensitive attributes. In this paper, we present a novel technique called slicing, which partitions the data both horizontally and vertically. We show that slicing preserves better data utility than generalization and can be used for membership disclosure protection. Another important advantage of slicing is that it can handle high-dimensional data. We show how slicing can be used for attribute disclosure protection and develop an efficient algorithm for computing the sliced data that obey the ℓ-diversity requirement. Our workload experiments confirm that slicing preserves better utility than generalization and is more effective than bucketization in workloads involving the sensitive attribute. Our experiments also demonstrate that slicing can be used to prevent membership disclosure.

Most existing Multiple-Instance Learning (MIL) algorithms assume data instances and/or data bags are independently and identically distributed. But there often exists rich additional dependency/structure information between instances/bags within many applications of MIL. Ignoring this structure information limits the performance of existing MIL algorithms. This paper explores the research problem as multiple instance learning on structured data (MILSD) and formulates a novel framework that considers additional structure information. In particular, an effective and efficient optimization algorithm has been proposed to solve the original non-convex optimization problem by using a combination of Concave-Convex Constraint Programming (CCCP) method and an adapted Cutting Plane method, which deals with two sets of constraints caused by learning on instances within individual bags and learning on structured data. Our method has the nice convergence property, with specified precision on each set of constraints. Experimental results on three different applications, i.e., webpage classification, market targeting, and protein fold identification, clearly demonstrate the advantages of the proposed method over state-of-the-art methods.

Recent work has shown the importance of considering the adversary's background knowledge when reasoning about privacy in data publishing. However, it is very difficult for the data publisher to know exactly the adversary's background knowledge. Existing work cannot satisfactorily model background knowledge and reason about privacy in the presence of such knowledge. This paper presents a general framework for modeling the adversary's background knowledge using kernel estimation methods. This framework subsumes different types of knowledge (e.g., negative association rules) that can be mined from the data. Under this framework, we reason about privacy using Bayesian inference techniques and propose the skyline (B, t)-privacy model, which allows the data publisher to enforce privacy requirements to protect the data against adversaries with different levels of background knowledge. Through an extensive set of experiments, we show the effects of probabilistic background knowledge in data anonymization and the effectiveness of our approach in both privacy protection and utility preservation.

We develop a cyclical blockwise coordinate descent algorithm for the multi-task Lasso that efficiently solves problems with thousands of features and tasks. The main result shows that a closed-form Winsorization operator can be obtained for the sup-norm penalized least squares regression. This allows the algorithm to find solutions to very large-scale problems far more efficiently than existing methods. This result complements the pioneering work of Friedman, et al. (2007) for the single-task Lasso. As a case study, we use the multi-task Lasso as a variable selector to discover a semantic basis for predicting human neural activation. The learned solution outperforms the standard basis for this task on the majority of test participants, while requiring far fewer assumptions about cognitive neuroscience. We demonstrate how this learned basis can yield insights into how the brain represents the meanings of words.

In this paper we consider the problem of classification in the presence of pairwise constraints, which consist of pairs of examples as well as a binary variable indicating whether they belong to the same class or not. We propose a method which can effectively utilize pairwise constraints to construct an estimator of the decision boundary, and we show that the resulting estimator is sign-insensitive consistent with respect to the optimal linear decision boundary. We also study the asymptotic variance of the estimator and extend the method to handle both labeled and pairwise examples in a natural way. Several experiments on simulated datasets and real world classification datasets are conducted. The results not only verify the theoretical properties of the proposed method but also demonstrate its practical value in applications. 1.

To deal with the problem of insufficient labeled data in video object classification, one solution is to utilize additional pairwise constraints that indicate the relationship between two examples, i.e., whether these examples belong to the same class or not. In this paper, we propose a discriminative learning approach which can incorporate pairwise constraints into a conventional margin-based learning framework. Different from previous work that usually attempts to learn better distance metrics or estimate the underlying data distribution, the proposed approach can directly model the decision boundary and, thus, require fewer model assumptions. Moreover, the proposed approach can handle both labeled data and pairwise constraints in a unified framework. In this work, we investigate two families of pairwise loss functions, namely, convex and nonconvex pairwise loss functions, and then derive three pairwise learning algorithms by plugging in the hinge loss and the logistic loss functions. The proposed learning algorithms were evaluated using a people identification task on two surveillance video data sets. The experiments demonstrated that the proposed pairwise learning algorithms considerably outperform the baseline classifiers using only labeled data and two other pairwise learning algorithms with the same amount of pairwise constraints.

AdaBoost.OC has shown to be an effective method in boosting "weak" binary classifiers for multi-class learning. It employs the Error Correcting Output Code (ECOC) method to convert a multi-class learning problem into a set of binary classification problems, and applies the AdaBoost algorithm to solve them efficiently. In this paper, we propose a new boosting algorithm that improves the AdaBoost.OC algorithm in two aspects: 1) It introduces a smoothing mechanism into the boosting algorithm to alleviate the potential overfitting problem with the AdaBoost algorithm, and 2) It introduces a probabilistic coding scheme to generate binary codes for multiple classes such that training errors can be efficiently reduced. Empirical studies with seven UCI datasets have indicated that the proposed boosting algorithm is more robust and effective than the AdaBoost.OC algorithm for multi-class learning.

We propose a probabilistic model based on Independent Component Analysis for learning multiple related tasks. In our model the task parameters are assumed to be generated from independent sources which account for the relatedness of the tasks. We use Laplace distributions to model hidden sources which makes it possible to identify the hidden, independent components instead of just modeling correlations. Furthermore, our model enjoys a sparsity property which makes it both parsimonious and robust. We also propose efficient algorithms for both empirical Bayes method and point estimation. Our experimental results on two multi-label text classification data sets show that the proposed approach is promising.

This paper reports a cross-benchmark evaluation of regularized logistic regression (LR) and incremental Rocchio for adaptive filtering. Using four corpora from the Topic Detection and Tracking (TDT) forum and the Text Retrieval Conferences (TREC) we evaluated these methods with non-stationary topics at various granularity levels, and measured performance with different utility settings. We found that LR performs strongly and robustly in optimizing T11SU (a TREC utility function) while Rocchio is better for optimizing Ctrk (the TDT tracking cost), a high-recall oriented objective function. Using systematic cross-corpus parameter optimization with both methods, we obtained the best results ever reported on TDT5, TREC10 and TREC11. Relevance feedback on a small portion (0.05~0.2%) of the TDT5 test documents yielded significant performance improvements, measuring up to a 54% reduction in Ctrk and a 20.9% increase in T11SU (with b=0.1), compared to the results of the top-performing system in TDT2004 without relevance feedback information.

Well-calibrated probabilities are necessary in many applications like probabilistic frameworks or cost-sensitive tasks. Based on previous success of asymmetric Laplace method in calibrating text classifiers' scores, we propose to use piecewise logistic regression, which is a simple extension of standard logistic regression, as an alternative method in the discriminative family. We show that both methods have the flexibility to be piecewise linear functions in log-odds, but they are based on quite different assumptions. We evaluated asymmetric Laplace method, piecewise logistic regression and standard logistic regression over standard text categorization collections (Reuters-21578 and TRECAP) with three classifiers (SVM, Naive Bayes and Logistic Regression Classifier), and observed that piecewise logistic regression performs significantly better than the other two methods in the log-loss metric.

In this paper we propose a probabilistic model for online document clus- tering. We use non-parametric Dirichlet process prior to model the grow- ing number of clusters, and use a prior of general English language model as the base distribution to handle the generation of novel clusters. Furthermore, cluster uncertainty is modeled with a Bayesian Dirichlet- multinomial distribution. We use empirical Bayes method to estimate hyperparameters based on a historical dataset. Our probabilistic model is applied to the novelty detection task in Topic Detection and Tracking (TDT) and compared with existing approaches in the literature.

Real-world applications of text categorization often require a system to deal with tens of thousands of categories de-fined over a large taxonomy. This paper addresses the prob-lem with respect to a set of popular algorithms in text cat-egorization, including Support Vector Machines, k-nearest neighbor, ridge regression, linear least square fit and logistic regression. By providing a formal analysis of the compu-tational complexity of each classification method, followed by an investigation on the usage of different classifiers in a hierarchical setting of categorization, we show how the scalability of a method depends on the topology of the hi-erarchy and the category distributions. In addition, we are able to obtain tight bounds for the complexities by using the power law to approximate category distributions over a hi-erarchy. Experiments with kNN and SVM classifiers on the OHSUMED corpus are reported on, as concrete examples.

Real-world applications often require the classification of documents under situations of small number of features, mis-labeled documents and rare positive examples. This paper investigates the robustness of three regularized linear classification methods (SVM, ridge regression and logistic regression) under above situations. We compare these methods in terms of their loss functions and score distributions, and establish the connection between their optimization problems and generalization error bounds. Several sets of controlled experiments on the Reuters-21578 corpus are conducted to investigate the robustness of these methods. Our results show that ridge regression seems to be the most promising candidate for rare class problems.

Automated detection of the first document reporting each new event in temporally-sequenced streams of documents is an open challenge. In this paper we propose a new approach which addresses this problem in two stages: 1) using a supervised learning algorithm to classify the on-line document stream into pre-defined broad topic categories, and 2) performing topic-conditioned novelty detection for documents in each topic. We also focus on exploiting named-entities for event-level novelty detection and using feature-based heuristics derived from the topic histories. Evaluating these methods using a set of broadcast news stories, our results show substantial performance gains over the traditional one-level approach to the novelty detection problem.