Katrin Erk

VECTOR SPACE MODELS OF WORD MEANING AND PHRASE MEANING: A SURVEY

Distributional models represent a word through the contexts in which it has been observed. They can be used to predict similarity in meaning, based on the distributional hypothesis, which states that two words that occur in similar contexts tend to have similar meanings. Distributional approaches are often implemented in vector space models. They represent a word as a point in high-dimensional space, where each dimension stands for a context item, and a words coordinates represent its context counts. Occurrence in similar contexts then means proximity in space. In this survey we look at the use of vector space models to describe the meaning of words and phrases: the phenomena that vector space models address, and the techniques that they use to do so. Many word meaning phenomena can be described in terms of semantic similarity: synonymy, priming, categorization, and the typicality of a predicates arguments. But vector space models can do more than just predict semantic similarity. They are a very flexible tool, because they can make use of all of linear algebra, with all its data structures and operations. The dimensions of a vector space can stand for many things: context words, or non-linguistic context like images, or properties of a concept. And vector space models can use matrices or higher-order arrays instead of vectors for representing more complex relationships. Polysemy is a tough problem for distributional approaches, as a representation that is learned from all of a words contexts will conflate the different senses of the word. It can be addressed, using either clustering or vector combination techniques. Finally, we look at vector space models for phrases, which are usually constructed by combining word vectors. Vector space models for phrases can predict phrase similarity, and some argue that they can form the basis for a general-purpose representation framework for natural language semantics.

SemEval-2007 Task 19: Frame Semantic Structure Extraction

This task consists of recognizing words and phrases that evoke semantic frames as defined in the FrameNet project (http://framenet.icsi.berkeley.edu), and their semantic dependents, which are usually, but not always, their syntactic dependents (including subjects). The training data was FN annotated sentences. In testing, participants automatically annotated three previously unseen texts to match gold standard (human) annotation, including predicting previously unseen frames and roles. Precision and recall were measured both for matching of labels of frames and FEs and for matching of semantic dependency trees based on the annotation.

Towards a Resource for Lexical Semantics: A Large German Corpus with Extensive Semantic Annotation

We describe the ongoing construction of a large, semantically annotated corpus resource as reliable basis for the large-scale acquisition of word-semantic information, e.g. the construction of domain-independent lexica. The backbone of the annotation are semantic roles in the frame semantics paradigm. We report experiences and evaluate the annotated data from the first project stage. On this basis, we discuss the problems of vagueness and ambiguity in semantic annotation.

A flexible, corpus-driven model of regular and inverse selectional preferences

We present a vector space–based model for selectional preferences that predicts plausibility scores for argument headwords. It does not require any lexical resources (such as WordNet). It can be trained either on one corpus with syntactic annotation, or on a combination of a small semantically annotated primary corpus and a large, syntactically analyzed generalization corpus. Our model is able to predict inverse selectional preferences, that is, plausibility scores for predicates given argument heads. We evaluate our model on one NLP task (pseudo-disambiguation) and one cognitive task (prediction of human plausibility judgments), gauging the influence of different parameters and comparing our model against other model classes. We obtain consistent benefits from using the disambiguation and semantic role information provided by a semantically tagged primary corpus. As for parameters, we identify settings that yield good performance across a range of experimental conditions. However, frequency remains a major influence of prediction quality, and we also identify more robust parameter settings suitable for applications with many infrequent items.

Graded Word Sense Assignment

Word sense disambiguation is typically phrased as the task of labeling a word in context with the best-fitting sense from a sense inventory such as WordNet. While questions have often been raised over the choice of sense inventory, computational linguists have readily accepted the best-fitting sense methodology despite the fact that the case for discrete sense boundaries is widely disputed by lexical semantics researchers. This paper studies graded word sense assignment, based on a recent dataset of graded word sense annotation.

Probabilistic Soft Logic for Semantic Textual Similarity

Probabilistic Soft Logic (PSL) is a recently developed framework for probabilistic logic. We use PSL to combine logical and distributional representations of natural-language meaning, where distributional information is represented in the form of weighted inference rules. We apply this framework to the task of Semantic Textual Similarity (STS) (i.e. judging the semantic similarity of naturallanguage sentences), and show that PSL gives improved results compared to a previous approach based on Markov Logic Networks (MLNs) and a purely distributional approach.

Representing words as regions in vector space

Vector space models of word meaning typically represent the meaning of a word as a vector computed by summing over all its corpus occurrences. Words close to this point in space can be assumed to be similar to it in meaning. But how far around this point does the region of similar meaning extend? In this paper we discuss two models that represent word meaning as regions in vector space. Both representations can be computed from traditional point representations in vector space. We find that both models perform at over 95% F-score on a token classification task.

Measuring Word Meaning in Context

Word sense disambiguation (WSD) is an old and important task in computational linguistics that still remains challenging, to machines as well as to human annotators. Recently there have been several proposals for representing word meaning in context that diverge from the traditional use of a single best sense for each occurrence. They represent word meaning in context through multiple paraphrases, as points in vector space, or as distributions over latent senses. New methods of evaluating and comparing these different representations are needed.In this paper we propose two novel annotation schemes that characterize word meaning in context in a graded fashion. In WSsim annotation, the applicability of each dictionary sense is rated on an ordinal scale. Usim annotation directly rates the similarity of pairs of usages of the same lemma, again on a scale. We find that the novel annotation schemes show good inter-annotator agreement, as well as a strong correlation with traditional single-sense annotation and with annotation of multiple lexical paraphrases. Annotators make use of the whole ordinal scale, and give very fine-grained judgments that “mix and match” senses for each individual usage. We also find that the Usim ratings obey the triangle inequality, justifying models that treat usage similarity as metric.There has recently been much work on grouping senses into coarse-grained groups. We demonstrate that graded WSsim and Usim ratings can be used to analyze existing coarse-grained sense groupings to identify sense groups that may not match intuitions of untrained native speakers. In the course of the comparison, we also show that the WSsim ratings are not subsumed by any static sense grouping.

What Substitutes Tell Us - Analysis of an "All-Words" Lexical Substitution Corpus

We present the first large-scale English “allwords lexical substitution” corpus. The size of the corpus provides a rich resource for investigations into word meaning. We investigate the nature of lexical substitute sets, comparing them to WordNet synsets. We find them to be consistent with, but more fine-grained than, synsets. We also identify significant differences to results for paraphrase ranking in context reported for the SEMEVAL lexical substitution data. This highlights the influence of corpus construction approaches on evaluation results.

Measuring semantic relatedness with vector space models and random walks

Both vector space models and graph random walk models can be used to determine similarity between concepts. Noting that vectors can be regarded as local views of a graph, we directly compare vector space models and graph random walk models on standard tasks of predicting human similarity ratings, concept categorization, and semantic priming, varying the size of the dataset from which vector space and graph are extracted.