Distributional Models of Category Concepts Based on Names of Category Members

Westera, Matthijs; Gupta, Abhijeet; Boleda, Gemma; Padó, Sebastian

doi:10.1111/cogs.13029

Cited by 6 publications

(6 citation statements)

References 79 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A different question is why should categories of people be harder to decode than categories of places in classification analyses ( Figure 7 ). We do not have a full story here, but we will just point out that social concepts are an idiosyncratic type of category, falling in between abstract and concrete concepts (Anderson et al, 2014 ; Rice et al, 2018 ; Conca et al, 2021 ), and that we found, consistently with the literature (Westera et al, 2021 ), a similar pattern of results in our clustering analysis for distributional models ( Figure 2 ).…”

Section: Discussionsupporting

confidence: 88%

“…Performance is above chance for most models in all possible labellings, for fine-grained categorical labels as well. The toughest discrimination is the one where both individual and category vectors for people are used, and in this respect others have already found that social categories for people are not well captured by distributional models (Westera et al, 2021 ). Also, in general, performance worsens when clustering vectors for individuals and categories together, as already observed by Gupta et al ( 2018 ).…”

Section: Methodsmentioning

confidence: 99%

“…We assume that the core result patterns would translate also to the other models in the family. In this respect, the most similar approach to ours is that of Westera et al (2021), where the authors show, by using just one model, Word2Vec, that better representations for categories of entities can be obtained by averaging exemplars instead of acquiring separate vectors for the categories, by evaluating the vectors on a set of human judgments.…”

Section: Decoding Individual Entities Using Word Vectorsmentioning

confidence: 95%

See 2 more Smart Citations

Exploring the Representations of Individual Entities in the Brain Combining EEG and Distributional Semantics

Bruera

Poesio

2022

Front. Artif. Intell.

View full text Add to dashboard Cite

Semantic knowledge about individual entities (i.e., the referents of proper names such as Jacinta Ardern) is fine-grained, episodic, and strongly social in nature, when compared with knowledge about generic entities (the referents of common nouns such as politician). We investigate the semantic representations of individual entities in the brain; and for the first time we approach this question using both neural data, in the form of newly-acquired EEG data, and distributional models of word meaning, employing them to isolate semantic information regarding individual entities in the brain. We ran two sets of analyses. The first set of analyses is only concerned with the evoked responses to individual entities and their categories. We find that it is possible to classify them according to both their coarse and their fine-grained category at appropriate timepoints, but that it is hard to map representational information learned from individuals to their categories. In the second set of analyses, we learn to decode from evoked responses to distributional word vectors. These results indicate that such a mapping can be learnt successfully: this counts not only as a demonstration that representations of individuals can be discriminated in EEG responses, but also as a first brain-based validation of distributional semantic models as representations of individual entities. Finally, in-depth analyses of the decoder performance provide additional evidence that the referents of proper names and categories have little in common when it comes to their representation in the brain.

show abstract

Section: Discussionsupporting

confidence: 88%

Section: Methodsmentioning

confidence: 99%

Section: Decoding Individual Entities Using Word Vectorsmentioning

confidence: 95%

See 1 more Smart Citation

Exploring the Representations of Individual Entities in the Brain Combining EEG and Distributional Semantics

Bruera

Poesio

2022

Front. Artif. Intell.

View full text Add to dashboard Cite

show abstract

“…Mack, Love, and Preston (2018) observed that multiple episodes must be experienced in the process of concept formation. Westera, Gupta, Boleda, and Padó (2021) demonstrated that concept categories can be formed through exposure to vocabulary in a bottom-up manner, with high frequency of temporally and conceptually co-occuring words forming a stronger part of the mental network of the concept than terms less frequently associated (for further evidence, see Banks, Wingfield, & Connell, 2021; Mack et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

The Aesthetic Emotion Lexicon: A Literature Review of Emotion Words Used by Researchers to Describe Aesthetic Experiences

Schubert

2023

Empirical Studies of the Arts

View full text Add to dashboard Cite

The emotions that can be considered members of the set of Aesthetic Emotions (AEs) is controversial. The present study investigated the terms used by researchers in peer reviewed studies to exemplify AEs. 100 publications from 2000–2019 exemplifying AE terms were located to produced 111 AEs which were proposed as the basis of an AE lexicon. Awe, (being) moved and wonder were reliable members and without contradiction. One fifth were negatively valenced (e.g., anger, disgust), suggesting that the presence of negative AEs is generally accepted but not reliably. One quarter of the entries were also non-AEs and an additional 20 were exclusively so, producing a total of 131 terms. The lexicon is a concrete, dynamic set of examples against which to investigate extant definitions of AEs and to further develop theory. The robust presence of three terms suggests that calls to abandon the concept of AE may be premature.

show abstract

“…The relationship between word representations and concepts is nuanced; seeWestera & Boleda 2019;Westera et al 2021 for discussion.…”

mentioning

confidence: 99%

What’s in a name? A large-scale computational study on how competition between names affects naming variation

Gualdoni¹,

Mädebach²,

Brochhagen³

et al. 2022

Preprint

View full text Add to dashboard Cite

Speakers often use different names to refer to the same entity (e.g., “woman” vs.“tennis player”). We study how typicality affects variation in naming visually presented objects. We use a novel computational approach to estimate visual typicality from images, and analyze a large dataset containing naming data for realistic images. In contrast to previous work, we take into account the visual properties of both the object and the scene in which it appears; and factor in multiple candidate names. We show that visual typicality mediates competition between candidate names: high competition, induced by the relationship between the visual properties of the object and the visual representations associated to names, predicts higher naming variation. On a methodological level, we demonstrate the potential of using large-scale datasets with realistic images in conjunction with computational methods to shed light on how people name objects.

show abstract

Distributional Models of Category Concepts Based on Names of Category Members

Cited by 6 publications

References 79 publications

Exploring the Representations of Individual Entities in the Brain Combining EEG and Distributional Semantics

Exploring the Representations of Individual Entities in the Brain Combining EEG and Distributional Semantics

The Aesthetic Emotion Lexicon: A Literature Review of Emotion Words Used by Researchers to Describe Aesthetic Experiences

What’s in a name? A large-scale computational study on how competition between names affects naming variation

Contact Info

Product

Resources

About