Probabilistic Latent Semantic Analysis is a novel statistical technique for the analysis of two--mode and co-occurrence data, which has applications in information retrieval and filtering, natural language processing, machine learning from text, and in related areas. Compared to standard Latent Semantic Analysis which stems from linear algebra and performs a Singular Value Decomposition of co-occurrence tables, the proposed method is based on a mixture decomposition derived from a latent class model. This results in a more principled approach which has a solid foundation in statistics. In order to avoid overfitting, we propose a widely applicable generalization of maximum likelihood model fitting by tempered EM. Our approach yields substantial and consistent improvements over Latent Semantic Analysis in a number of experiments.

Abstract. This paper presents a novel statistical method for factor analysis of binary and count data which is closely related to a technique known as Latent Semantic Analysis. In contrast to the latter method which stems from linear algebra and performs a Singular Value Decomposition of co-occurrence tables, the proposed technique uses a generative latent class model to perform a probabilistic mixture decomposition. This results in a more principled approach with a solid foundation in statistical inference. More precisely, we propose to make use of a temperature controlled version of the Expectation Maximization algorithm for model fitting, which has shown excellent performance in practice. Probabilistic Latent Semantic Analysis has many applications, most prominently in information retrieval, natural language processing, machine learning from text, and in related areas. The paper presents perplexity results for different types of text and linguistic data collections and discusses an application in automated document indexing. The experiments indicate substantial and consistent improvements of the probabilistic method over standard Latent Semantic Analysis.

Statistical approaches to language learning typically focus on either short-range syntactic dependencies or long-range semantic dependencies between words. We present a generative model that uses both kinds of dependencies, and can be used to simultaneously find syntactic classes and semantic topics despite having no representation of syntax or semantics beyond statistical dependency. This model is competitive on tasks like part-of-speech tagging and document classification with models that exclusively use short- and long-range dependencies respectively. 1

Image auto-annotation, i.e., the association of words to whole images, has attracted considerable attention. In particular, unsupervised, probabilistic latent variable models of text and image features have shown encouraging results, but their performance with respect to other approaches remains unknown. In this paper, we apply and compare two simple latent space models commonly used in text analysis, namely Latent Semantic Analysis (LSA) and Probabilistic LSA (PLSA). Annotation strategies for each model are discussed. Remarkably, we found that, on a 8000-image dataset, a classic LSA model defined on keywords and a very basic image representation performed as well as much more complex, state-of-the-art methods. Furthermore, nonprobabilistic methods (LSA and direct image matching) outperformed PLSA on the same dataset.

This paper proposes a framework for representing the meaning of phrases and sentences in vector space. Central to our approach is vector composition which we operationalize in terms of additive and multiplicative functions. Under this framework, we introduce a wide range of composition models which we evaluate empirically on a sentence similarity task. Experimental results demonstrate that the multiplicative models are superior to the additive alternatives when compared against human judgments.

In this paper, we propose a novel statistical language model to capture topic-related long-range dependencies. Topics are modeled in a latent variable framework in which we also derive an EM algorithm to perform a topic factor decomposition based on a segmented training corpus. The topic model is combined with a standard language model to be used for on-line word prediction. Perplexity results indicate an improvement over previously proposed topic models, which unfortunately has not translated into lower word error.

Speech recognition performance is severely affected when the lexical, syntactic, or semantic characteristics of the discourse in the training and recognition tasks differ. The aim of language model adaptation is to exploit specific, albeit limited, knowledge about the recognition task to compensate for this mismatch. More generally, an adaptive language model seeks to maintain an adequate representation of the current task domain under changing conditions involving potential variations in vocabulary, syntax, content, and style. This paper presents an overview of the major approaches proposed to address this issue, and offers some perspectives regarding their comparative merits and associated tradeoffs. Ó 2003 Elsevier B.V. All rights reserved. 1.

In-domain texts for estimating statistical language models are not easily found for most languages of the world. We present two techniques to take advantage of in-domain text resources in other languages. First, we extend the notion of lexical triggers, which have been used monolingually for language model adaptation, to the cross-lingual problem, permitting the construction of sharper language models for a target-language document by drawing statistics from related documents in a resource-rich language. Next, we show that cross-lingual latent semantic analysis is similarly capable of extracting useful statistics for language modeling. Neither technique requires explicit translation capabilities between the two languages! We demonstrate significant reductions in both perplexity and word error rate on a Mandarin speech recognition task by using these techniques.

We use a probabilistic mixture decomposition method to determine topics in the Pennsylvania Gazette, a major colonial U.S. newspaper from 1728–1800. We assess the value of several topic decomposition techniques for historical research and compare the accuracy and efficacy of various methods. After determining the topics covered by the 80,000 articles and advertisements in the entire 18th century run of the Gazette, we calculate how the prevalence of those topics changed over time, and give historically relevant examples of our findings. This approach reveals important information about the content of this colonial newspaper, and suggests the value of such approaches to a more complete understanding of early American print culture and society.

In this paper we propose a novel statistical language model to capture long-range semantic dependencies. Specifically, we apply the concept of semantic composition to the problem of constructing predictive history representations for upcoming words. We also examine the influence of the underlying semantic space on the composition task by comparing spatial semantic representations against topic-based ones. The composition models yield reductions in perplexity when combined with a standard n-gram language model over the n-gram model alone. We also obtain perplexity reductions when integrating our models with a structured language model. 1