Traditional wisdom holds that once documents are turned into bag-of-words (unigram count) vectors, word orders are completely lost. We introduce an approach that, perhaps surprisingly, is able to learn a bigram language model from a set of bag-of-words documents. At its heart, our approach is an EM algorithm that seeks a model which maximizes the regularized marginal likelihood of the bagof-wor...

Cross-language learning allows one to use training data from one language to build models for a different language. Many approaches to bilingual learning require that we have word-level alignment of sentences from parallel corpora. In this work we explore the use of autoencoder-based methods for cross-language learning of vectorial word representations that are coherent between two languages, w...

Traditional information retrieval systems use query words to identify relevant documents. In difficult retrieval tasks, however, one needs access to a wealth of background knowledge. We present a method that uses Wikipedia-based feature generation to improve retrieval performance. Intuitively, we expect that using extensive world knowledge is likely to improve recall but may adversely affect pr...

In this paper, we present an approach based on probabilistic latent semantic analysis (PLSA) to achieve the task of automatic image annotation and retrieval. In order to model training data precisely, each image is represented as a bag of visual words. Then a probabilistic framework is designed to capture semantic aspects from visual and textual modalities, respectively. Furthermore, an adaptiv...

This article describes the implementation of I2R word sense disambiguation system (I2R −WSD) that participated in one senseval3 task: Chinese lexical sample task. Our core algorithm is a supervised Naive Bayes classifier. This classifier utilizes an optimal feature set, which is determined by maximizing the cross validated accuracy of NB classifier on training data. The optimal feature set incl...

Most text categorization systems use simple models of documents and document collections. In this paper we describe a technique that improves a simple web page classifier’s performance on pages from a new, unseen web site, by exploiting link structure within a site as well as page structure within hub pages. On real-world test cases, this technique significantly and substantially improves the a...

In two eye-tracking experiments, the effects of salience in accent training and speech accentedness on spoken-word recognition were investigated. Salience was expected to increase a stimulus' prominence and therefore promote learning. A training-test paradigm was used on native German participants utilizing an artificial German accent. Salience was elicited by two different criteria: production...

This paper presents the results of the RepEval 2017 Shared Task, which evaluated neural network sentence representation learning models on the MultiGenre Natural Language Inference corpus (MultiNLI) recently introduced by Williams et al. (2017). All of the five participating teams beat the bidirectional LSTM (BiLSTM) and continuous bag of words baselines reported in Williams et al.. The best si...

In this article, we examine the effectiveness of bootstrapping supervised machine-learning polarity classifiers with the help of a domain-independent rulebased classifier that relies on a lexical resource, i.e., a polarity lexicon and a set of linguistic rules. The benefit of this method is that though no labeled training data are required, it allows a classifier to capture in-domain knowledge ...