Human EEG recordings for 1,854 concepts presented in rapid serial visual presentation streams

The neural basis of object recognition and semantic knowledge has been extensively studied but the high dimensionality of object space makes it challenging to develop overarching theories on how the brain organises object knowledge. To help understand how the brain allows us to recognise, categorise, and represent objects and object categories, there is a growing interest in using large-scale image databases for neuroimaging experiments. In the current paper, we present THINGS-EEG, a dataset containing human electroencephalography responses from 50 subjects to 1,854 object concepts and 22,248 images in the THINGS stimulus set, a manually curated and high-quality image database that was specifically designed for studying human vision. The THINGS-EEG dataset provides neuroimaging recordings to a systematic collection of objects and concepts and can therefore support a wide array of research to understand visual object processing in the human brain.

A lexical semantic approach to the L2 acquisition of Spanish psych verbs

This study builds on prior research on second language (L2) Spanish psych verbs, which has centered on morphosyntactic properties, by examining their syntactic distribution, which relies on lexical semantic knowledge. The fact that certain forms are licensed for some verbs, but not others, is the result of an underlying lexical semantic difference across verb classes, represented here as a difference in formal feature strength. To fully acquire the relevant grammatical distribution, L2 learners must successfully acquire (i) licensing restrictions on argument structure and (ii) underlying lexical semantic representations of individual verbs. Three groups of L2 learners ( n = 66) and a group of native Spanish speakers ( n = 19) completed two judgment tasks (one with aural stimuli and one with written stimuli) which presented object experiencer psych verbs in multiple argument structures. Results show that advanced L2 learners are largely sensitive to the distribution tested here; however, while they have acquired relevant licensing restrictions, they may associate fixed feature settings with verbs that allow variable feature settings. These results are consistent with predictions made by the Feature Reassembly Hypothesis and highlight the role of lexical semantic features in second language acquisition.

Context-Fused Guidance for Image Captioning Using Sequence-Level Training

Recent image captioning models based on the encoder-decoder framework have achieved remarkable success in humanlike sentence generation. However, an explicit separation between encoder and decoder brings out a disconnection between the image and sentence. It usually leads to a rough image description: the generated caption only contains main instances but neglects additional objects and scenes unexpectedly, which reduces the caption consistency of the image. To address this issue, we proposed an image captioning system within context-fused guidance in this paper. It incorporates regional and global image representation as the compositional visual features to learn the objects and attributes in images. To integrate image-level semantic information, the visual concept is employed. To avoid misleading decoding, a context fusion gate is introduced to calculate the textual context by selectively aggregating the information of visual concept and word embedding. Subsequently, the context-fused image guidance is formulated based on the compositional visual features and textual context. It provides the decoder with informative semantic knowledge. Finally, a captioner with a two-layer LSTM architecture is constructed to generate captions. Moreover, to overcome the exposure bias, we train the proposed model through sequence decision-making. The experiments conducted on the MS COCO dataset show the outstanding performance of our work. The linguistic analysis demonstrates that our model improves the caption consistency of the image.

Semantic City Planning Systems (SCPS): A Literature Review

This review focuses on recent research literature on the use of Semantic Web Technologies (SWT) in city planning. The review foregrounds representational, evaluative, projective, and synthetical meta-practices as constituent practices of city planning. We structure our review around these four meta-practices that we consider fundamental to those processes. We find that significant research exists in all four metapractices. Linking across domains by combining various methods of semantic knowledge generation, processing, and management is necessary to bridge gaps between these meta-practices and will enable future Semantic City Planning Systems.

The Role of the Angular Gyrus in Semantic Cognition – A Synthesis of Five Functional Neuroimaging Studies

Semantic knowledge is central to human cognition. The angular gyrus (AG) is widely considered a key brain region for semantic cognition. However, the role of the AG in semantic processing is controversial. Key controversies concern response polarity (activation vs. deactivation) and its relation to task difficulty, lateralization (left vs. right AG), and functional-anatomical subdivision (PGa vs. PGp subregions). Here, we combined the fMRI data of five studies on semantic processing (n = 172) and analyzed the response profiles from the same anatomical regions-of-interest for left and right PGa and PGp. We found that the AG was consistently deactivated during non-semantic conditions, whereas response polarity during semantic conditions was inconsistent. However, the AG consistently showed relative response differences between semantic and non-semantic conditions, and between different semantic conditions. A combined analysis across all studies revealed that AG responses could be best explained by independent effects of both task difficulty and semantic processing demand. Task difficulty effects were stronger in PGa than PGp, regardless of hemisphere. Semantic effects were stronger in left than right AG, regardless of subregion. These results suggest that the AG is independently engaged in both domain-general task-difficulty-related processes and domain-specific semantic processes. In semantic processing, we propose that left AG acts as a “multimodal convergence zone” that binds different semantic features associated with the same concept, enabling efficient access to task-relevant features.

Completing WordNets with Sememe Knowledge

WordNets organize words into synonymous word sets, and the connections between words present the semantic relationships between them, which have become an indispensable source for natural language processing (NLP) tasks. With the development and evolution of languages, WordNets need to be constantly updated manually. To address the problem of inadequate word semantic knowledge of “new words”, this study explores a novel method to automatically update the WordNet knowledge base by incorporating word-embedding techniques with sememe knowledge from HowNet. The model first characterizes the relationships among words and sememes with a graph structure and jointly learns the embedding vectors of words and sememes; finally, it synthesizes word similarities to predict concepts (synonym sets) of new words. To examine the performance of the proposed model, a new dataset connected to sememe knowledge and WordNet is constructed. Experimental results show that the proposed model outperforms the existing baseline models.

The role of the angular gyrus in semantic cognition – A synthesis of five functional neuroimaging studies

Semantic knowledge is central to human cognition. The angular gyrus (AG) is widely considered a key brain region for semantic cognition. However, the role of the AG in semantic processing is controversial. Key controversies concern response polarity (activation vs. deactivation) and its relation to task difficulty, lateralization (left vs. right AG), and functional-anatomical subdivision (PGa vs. PGp subregions). Here, we combined the fMRI data of five studies on semantic processing (n = 172) and analyzed the response profiles from the same anatomical regions-of-interest for left and right PGa and PGp. We found that the AG was consistently deactivated during non-semantic conditions, whereas response polarity during semantic conditions was inconsistent. However, the AG consistently showed relative response differences between semantic and non-semantic conditions, and between different semantic conditions. A combined analysis across all studies revealed that AG responses could be best explained by independent effects of both task difficulty and semantic processing demand. Task difficulty effects were stronger in PGa than PGp, regardless of hemisphere. Semantic effects were stronger in left than right AG, regardless of subregion. These results suggest that the AG is independently engaged in both domain-general task-difficulty-related processes and domain-specific semantic processes. In semantic processing, we propose that left AG acts as a "multimodal convergence zone" that binds different semantic features associated with the same concept, enabling efficient access to task-relevant features.

Assessing lexical-semantic naming deficits using an automated machine learning approach

After left hemisphere stroke, 20-50% of people experience language deficits, including difficulties in naming. Naming errors that are semantically related to the intended target (e.g., producing “violin” for picture HARP) indicate a potential impairment in accessing knowledge of word forms and their meanings. Understanding the cause of naming impairments is crucial to better modeling of language production as well as for tailoring individualized rehabilitation. However, evaluation of naming errors is typically by subjective and laborious dichotomous classification. As a result, these evaluations do not capture the degree of semantic similarity and are susceptible to lower inter-rater reliability because of subjectivity. We investigated whether a computational linguistic measure, word2vec (Mikolov, Chen, Corrado, & Dean, 2013) addressed these limitations by evaluating errors during object naming in a group of patients during the acute stage of a left-hemisphere stroke (N=105). Pearson correlations demonstrated excellent convergent validity of word2vec’s semantically related estimates of naming errors and independent tests of access to lexical-semantic knowledge (p’s < .0001). Further, multiple regression analysis showed word2vec’s semantically related estimates were significantly better than human error classification at predicting performance on tests of lexical-semantic knowledge (p < .001). Useful to both theorists and clinicians, word2vec provides an automated, continuous, and objective psychometric measure of access to lexical-semantic knowledge during naming.

Did you see that? False memories for emotional words in bilingual children

When participants process a list of semantically strongly related words, the ones that were not presented may later be said, falsely, to have been on the list. This ‘false memory effect’ has been investigated by means of the DRM paradigm. We applied an emotional version of it to assess the false memory effect for emotional words in bilingual children with a minority language as L1 (their mother tongue) and a monolingual control group. We found that the higher emotionality of the words enhances memory distortion for both the bilingual and the monolingual children, in spite of the disadvantage related to vocabulary skills and of the socioeconomic status that acts on semantic processing independently from the condition of bilingualism. We conclude that bilingual children develop their semantic knowledge separately from their vocabulary skills and parallel to their monolingual peers, with a comparable role played by Arousal and Valence.