The Centre for Speech, Language and the Brain (CSLB) concept property norms

Devereux, Barry; Tyler, Lorraine K.; Geertzen, Jeroen; Randall, Billi

doi:10.3758/s13428-013-0420-4

Cited by 153 publications

(213 citation statements)

References 39 publications

Supporting

Mentioning

207

Contrasting

Unclassified

Order By: Relevance

“…Thus, the last model also included many features that were never presented in the experiment. The best approximation for multitude of semantic features associated with each target object was obtained by using a list of behaviorally produced object features from the Centre for Speech Language and the Brain dataset (henceforth, the CSLB features; Devereux et al, 2014). Using the same procedure as for the target words, we established a semantic coordinate for each clue and each of the newly listed features using the word2vec model a l l a v a i l a b l e f e a t u r e s Figure 2: Decoding performance using different models in the singletrial analyses.…”

Section: Semantic Representation Of the Target Object Is Best Decodedmentioning

confidence: 99%

“…cosine) between two concepts quantifies their semantic similarity. Semantic spaces can be obtained, for example by using statistical co-occurrence information collected from large text corpora (Erk, 2012;Kanerva and Ginter, 2014;Mikolov, Sutskever, et al, 2013;Turney and Pantel, 2010) or using behavioral methods to estimate similarity of descriptive content between items (Devereux et al, 2014;McRae, Cree, et al, 2005;Vinson and Vigliocco, 2008;Sudre et al, 2012). Such semantic spaces have been used as priors in machine learning based neural decoding models that have successfully associated various semantic feature sets (i.e.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Reconstructing meaning from bits of information

Kivisaari

Vliet

Hultén

et al. 2018

Preprint

View full text Add to dashboard Cite

We can easily identify a dog merely by the sound of barking or an orange by its citrus scent. In this work, we study the neural underpinnings of how the brain combines bits of information into meaningful object representations. Modern theories of semantics posit that the meaning of words can be decomposed into a unique combination of individual semantic features (e.g., "barks", "has citrus scent"). Here, participants received clues of individual objects in form of three isolated semantic features, given as verbal descriptions. We used machine-learning-based neural decoding to learn a mapping between individual semantic features and BOLD activation patterns. We discovered that the recorded brain patterns were best decoded using a combination of not only the three semantic features that were presented as clues, but a far richer set of semantic features typically linked to the target object. We conclude that our experimental protocol allowed us to observe how fragmented information is combined into a complete semantic representation of an object and suggest neuroanatomical underpinnings for this process.

show abstract

Section: Semantic Representation Of the Target Object Is Best Decodedmentioning

confidence: 99%

mentioning

confidence: 99%

Reconstructing meaning from bits of information

Kivisaari

Vliet

Hultén

et al. 2018

Preprint

View full text Add to dashboard Cite

show abstract

“…A wider range and more standardized set of concepts to choose from is provided by category norms (Battig & Montague, 1969;Van Overschelde et al, 2004) or object property norms (Devereux, Tyler, Geertzen, & Randall, 2014;McRae, Cree, Seidenberg, & McNorgan, 2005), which can be useful for improving comparability between studies. Despite their common use, it is important to note that the concepts and categories in those norms were selected mostly based on their use in previous studies.…”

Section: Selection Of Object Concepts and Object Categories In Previomentioning

confidence: 99%

“…Fourth, those categories could be used to generate a comprehensive set of typicality ratings of objects concepts. Finally, the concepts can form the basis for the creation of feature norms similar to existing ones (Devereux et al, 2014;McRae et al, 2005) or explicit ratings of object dimensions (e.g. real-world size, animacy, manipulability, etc.).…”

Section: Future Directionsmentioning

confidence: 99%

THINGS: A database of 1,854 object concepts and more than 26,000 naturalistic object images

Hebart

Dickter

Kidder

et al. 2019

Preprint

View full text Add to dashboard Cite

In recent years, the use of a large number of object concepts and naturalistic object images has been growing enormously in cognitive neuroscience research. Classical databases of object concepts are based mostly on a manually-curated set of concepts. Further, databases of naturalistic object images typically consist of single images of objects cropped from their background, or a large number of uncontrolled naturalistic images of varying quality, requiring elaborate manual image curation. Here we provide a set of 1,854 diverse object concepts sampled systematically from concrete picturable and nameable nouns in the American English language. Using these object concepts, we conducted a large-scale web image search to compile a database of 26,107 high-quality naturalistic images of those objects, with 12 or more object images per concept and all images cropped to square size. Using crowdsourcing, we provide higher-level category membership for the 27 most common categories and validate them by relating them to representations in a semantic embedding derived from large text corpora. Finally, by feeding images through a deep convolutional neural network, we demonstrate that they exhibit high selectivity for different object concepts, while at the same time preserving variability of different object images within each concept. Together, the THINGS database provides a rich resource of object concepts and object images and offers a tool for both systematic and large-scale naturalistic research in the fields of psychology, neuroscience, and computer science.

show abstract

“…For example, Devereux, Tyler, Geertzen, and Randall (2014) collected data for 638 concepts, thus extending the number of concepts selected by McRae and colleagues and including features that were produced by at least two participants (in McRae et al's, 2005, norms, the production frequency of a feature instead had to be greater than or equal to five for that feature to be included). Other recent norms have been collected by Lenci et al (2013), who included both blind and sighted participants.…”

Section: Existing Normsmentioning

confidence: 99%

Spanish semantic feature production norms for 400 concrete concepts

Vivas

Comesaña

et al. 2016

Behav Res

View full text Add to dashboard Cite

Semantic feature production norms provide many quantitative measures of different feature and concept variables that are necessary to solve some debates surrounding the nature of the organization, both normal and pathological, of semantic memory. Despite the current existence of norms for different languages, there are still no published norms in Spanish. This article presents a new set of norms collected from 810 participants for 400 living and nonliving concepts among Spanish speakers. These norms consist of empirical collections of features that participants used to describe the concepts. Four files were elaborated: a concept-feature file, a concept-concept matrix, a feature-feature matrix, and a significantly correlated features file. We expect that these norms will be useful for researchers in the fields of experimental psychology, neuropsychology, and psycholinguistics.

show abstract

The Centre for Speech, Language and the Brain (CSLB) concept property norms

Cited by 153 publications

References 39 publications

Reconstructing meaning from bits of information

Reconstructing meaning from bits of information

THINGS: A database of 1,854 object concepts and more than 26,000 naturalistic object images

Spanish semantic feature production norms for 400 concrete concepts

Contact Info

Product

Resources

About