Similarity of Semantic Relations

By Peter D. Turney. 2006

446 citations recorded by the source corpus.

There are at least two kinds of similarity. Relational similarity is correspondence between relations, in contrast with attributional similarity, which is correspondence between attributes. When two words have a high degree of attributional similarity, we call them synonyms. When two pairs of words have a high degree of relational similarity, we say that their relations are analogous. For example, the word pair mason:stone is analogous to the pair carpenter:wood. This article introduces Latent Relational Analysis (LRA), a method for measuring relational similarity. LRA has potential applications in many areas, including information extraction, word sense disambiguation, and information retrieval. Recently the Vector Space Model (VSM) of information retrieval has been adapted to measuring relational similarity, achieving a score of 47% on a collection of 374 college-level multiple-choice word analogy questions. In the VSM approach, the relation between a pair of words is characterized by a vector of frequencies of predefined patterns in a large corpus. LRA extends the VSM approach in three ways: (1) The patterns are derived automatically from the corpus, (2) the Singular Value Decomposition (SVD) is used to smooth the frequency data, and (3) automatically generated synonyms are used to explore variations of the word pairs. LRA achieves 56% on the 374 analogy questions, statistically equivalent to the average human score of 57%. On the related problem of classifying semantic relations, LRA achieves similar gains over the VSM.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Multi-Relational Latent Semantic Analysis (2013, background): “…role in applications for lexical semantics, such as word sense disambiguation (Boyd-Graber et al., 2007), measuring word ∗ Work conducted while interning at Microsoft Research. similarity (Deerwester et al., 1990) and relational similarity (Turney, 2006; Zhila et al., 2013; Mikolov et al., 2013).”
Improving Vector Space Word Representations Using Multilingual Correlation (2014, background): “…text (Deerwester et al., 1990; Turney and Pantel, 2010), 1 it is possible to construct high-quality semantic vectors — as judged by both correlations with human judgements of semantic relatedness (Turney, 2006; Agirre et al., 2009) and as features for downstream applications (Turian et al., 2010).”
code2vec: learning distributed representations of code (2018, methodology): “…paragraphs, and documents (such as “doc2vec”) [Le and Mikolov 2014] played a key role in unlocking the potential of neural networks for natural language processing (NLP) tasks [Bengio et al. 2003; Collobert and Weston 2008; Glorot et al. 2011; Socher et al. 2011; Turian et al. 2010; Turney 2006].”
Linguistic Regularities in Sparse and Explicit Word Representations (2014, methodology): “The word embeddings are designed to capture what Turney (2006) calls attributional similarities be-tween vocabulary items: words that appear in similar contexts will be close to each other in the projected space.”
Distributional Memory: A General Framework for Corpus-Based Semantics (2010, neutral): “Turney (2006b) deﬁnes relational similarity as the property shared by pairs of words (e.g, dog – animal and car – vehicle ) linked by similar semantic relations (e.g., hypernymy), despite the fact that the words in one pair might not be attributionally similar to those in the other pair (e.g., dog…”
Domain and Function: A Dual-Space Model of Semantic Relations and Compositions (2012, methodology): “The first set of experiments presents a dualspace model of semantic relations and evaluates the model with multiple choice analogy questions from the SAT.”
From 'F' to 'A' on the N.Y. Regents Science Exams: An Overview of the Aristo Project (2019, methodology): “Earlier work on standardized tests focused on specialized tasks, for example, SAT word analogies (Turney, 2006), GRE word antonyms (Mohammad et al., 2013), and TOEFL synonyms (Landauer and Dumais, 1997).”
A Comparison of Vector-based Representations for Semantic Composition (2012, methodology): “This ability has been put to practical use in numerous natural language processing tasks such as automatic thesaurus extraction (Grefenstette, 1994), word sense discrimination (Schütze, 1998), language modeling (Bellegarda, 2000), and the identification of analogical relations (Turney, 2006).”
SemEval-2012 Task 2: Measuring Degrees of Relational Similarity (2012, unknown): “, to express the same relation (Turney, 2006).”
Distributional Semantics Beyond Words: Supervised Learning of Analogy and Paraphrase (2013, methodology): “The simplest type of analogy is proportional analogy, which involves two pairs of words (Turney, 2006b).”
Deep Visual Analogy-Making (2015, methodology): “Analogies have been well-studied in the NLP community; Turney [28] used analogies from SAT tests to evaluate the performance of text analogy detection methods.”
Community Evaluation and Exchange of Word Vectors at wordvectors.org (2014, background): “, 1990; Turney and Pantel, 2010), it is possible to construct high-quality semantic vectors — as judged by both correlations with human judgements of semantic relatedness (Turney, 2006; Agirre et al., 2009) and as features for downstream applications (Turian et al.”
Combining Heterogeneous Models for Measuring Relational Similarity (2013, methodology): “Finally, we would like to evaluate the utility our approach in other applications, such as the SAT analogy problems proposed by Turney (2006) and question answering.”
How we BLESSed distributional semantic evaluation (2011, methodology): “Currently, distributional models for attributional similarity and relational similarity (Turney, 2006) are tested on different data sets, e.”
From Frequency to Meaning: Vector Space Models of Semantics (2010, methodology): “The idea of the VSM is to represent each document in a collection as a point in a space (a vector in a vector space).”
Distributional Models of Word Meaning (2018, unknown): “Turney PD. 2006.”
Wikipedia-based Semantic Interpretation for Natural Language Processing (2009, background): “stone are relationally similar, as the words in both pairs stand in the same relation (profession:material). State of the art results on relational similarity are based on Latent Relational Analysis (Turney, 2006, 2005). Sahami and Heilman (2006) proposed to use the Web as a source of additional kno”