Abstract. In many applications, it is necessary to algorithmically quantify the similarity exhibited by two strings composed of symbols from a finite alphabet. Numerous string similarity measures have been proposed. Particularly well-known measures are based are edit distance and the length of the longest common subsequence. We develop a notion of n-gram similarity and distance. We show that edit distance and the length of the longest common subsequence are special cases of n-gram distance and similarity, respectively. We provide formal, recursive definitions of n-gram similarity and distance, together with efficient algorithms for computing them. We formulate a family of word similarity measures based on n-grams, and report the results of experiments that suggest that the new measures outperform their unigram equivalents.

We approach the task of morphological inflection generation as discriminative string transduction. Our supervised system learns to generate word-forms from lemmas accompanied by morphological tags, and refines them by referring to the other forms within a paradigm.Results of experiments on six diverse languages with varying amounts of training data demonstrate that our approach improves the state of the art in terms of predicting inflected word-forms.

No abstract

We report results of experiments aimed at improving the translation quality by incorporating the cognate information into translation models. The results confirm that the cognate identification approach can improve the quality of word alignment in bitexts without the need for extra resources.

With the ever-growing popularity of online media such as blogs and social networking sites, the Internet is a valuable source of information for product and service reviews. Attempting to classify a subset of these documents using polarity metrics can be a daunting task. After a survey of previous research on sentiment polarity, we propose a novel approach based on Support Vector Machines. We compare our method to previously proposed lexical-based and machine learning (ML) approaches by applying it to a publicly available set of movie reviews. Our algorithm will be integrated within a blog visualization tool.

I present a method of identifying cognates in the vocabularies of related languages. I show that a measure of phonetic similarity based on multivalued features performs better than "orthographic" measures, such as the Longest Common Subsequence Ratio (LCSR) or Dice's coefficient. I introduce a procedure for estimating semantic similarity of glosses that employs keyword selection and WordNet. Tests performed on vocabularies of four Algonquian languages indicate that the method is capable of discovering on average nearly 75% percent of cognates at 50% precision.

Applying the noisy channel model to search query spelling correction requires an error model and a language model. Typically, the error model relies on a weighted string edit distance measure. The weights can be learned from pairs of misspelled words and their corrections. This paper investigates using the Expectation Maximization algorithm to learn edit distance weights directly from search query logs, without relying on a corpus of paired words.

Syllables play an important role in speech synthesis and recognition. We present several different approaches to the syllabification of phonemes. We investigate approaches based on linguistic theories of syllabification, as well as a discriminative learning technique that combines Support Vector Machine and Hidden Markov Model technologies. Our experiments on English, Dutch and German demonstrate that our transparent implementation of the sonority sequencing principle is more accurate than previous implementations, and that our language-independent SVM-based approach advances the current state-of-the-art, achieving word accuracy of over 98% in English and 99% in German and Dutch.