Unveiling the nature of interaction between semantics and phonology in lexical access based on multilayer networks

An essential aspect of human communication is the ability to access and retrieve information from ones’ ‘mental lexicon’. This lexical access activates phonological and semantic components of concepts, yet the question whether and how these two components relate to each other remains widely debated. We harness tools from network science to construct a large-scale linguistic multilayer network comprising of phonological and semantic layers. We find that the links in the two layers are highly similar to each other and that adding information from one layer to the other increases efficiency by decreasing the network overall distances, but specifically affecting shorter distances. Finally, we show how a multilayer architecture demonstrates the highest efficiency, and how this efficiency relates to weak semantic relations between cue words in the network. Thus, investigating the interaction between the layers and the unique benefit of a linguistic multilayer architecture allows us to quantify theoretical cognitive models of lexical access.

Neurophysiological correlates of disorder-related autobiographical memory in anorexia nervosa

Background Anorexia nervosa (AN) is characterized by an overgeneralization of food/body-related autobiographical memories (AM). This is regarded as an emotion regulation strategy with adverse long-term effects implicated in disorder maintenance and treatment resistance. Therefore, we aimed to examine neural correlates of food/body-related AM-recall in AN. Methods Twenty-nine female patients with AN and 30 medication-free age-sex-matched normal-weight healthy controls (HC) underwent functional magnetic resonance imaging while recalling AMs in response to food/body-related and neutral cue words. To control for general knowledge retrieval, participants engaged in a semantic generation and riser detection task. Results In comparison to HC, patients with AN generated fewer and less specific AMs in response to food/body-related words, but not for neutral cue words. Group comparisons revealed reduced activation in regions associated with self-referential processing and memory retrieval (precuneus and angular gyrus) during the retrieval of specific food/body-related AM in patients with AN. Brain connectivity in regions associated with memory functioning and executive control was reduced in patients with AN during the retrieval of specific food/body-related AM. Finally, resting-state functional connectivity analysis revealed no differences between groups, arguing against a general underlying disconnection of brain networks implicated in memory and emotional processing in AN. Conclusions These results indicate impaired neural processing of food/body-related AM in AN, with a reduced involvement of regions involved in self-referential processing. Our findings are discussed as possible neuronal correlates of emotional avoidance in AN and provide new insights of AN-pathophysiology underscoring the importance of targeting dysfunctional emotion regulation strategies during treatment.

A deep learning approach for identifying biomedical breakthrough discoveries using context analysis

Breakthrough research in scientific fields usually comes as a manifestation of major development and advancement. These advances build to an epiphany where new ways of thinking about a problem become possible. Identifying breakthrough research can be useful for cultivating and funding further innovation. This article presents a new method for identifying scientific breakthroughs from research papers based on cue words commonly associated with major advancements. We looked for specific terms signifying scientific breakthroughs in citing sentences to identify breakthrough articles. By setting a threshold for the number of citing sentences (“citances”) with breakthrough cue words that peer scholars often use when evaluating research, we identified articles containing breakthrough research. We call this approach the “others-evaluation” process. We then shortlisted candidates from the selected articles based on the authors’ evaluations of their own research, found in the abstracts. This we call the “self-evaluation” process. Combining the two approaches into a dual “others-self” evaluation process, we arrived at a sample of 237 potential breakthrough articles, most of which are recommended by the Faculty Opinions. Based on the breakthrough articles identified, using SVM, TextCNN, and BERT to train the models to identify abstracts with breakthrough evaluations. This automatic identification model can greatly simplify the process of others-self-evaluation process and promote identifying breakthrough research.

Cultural differences in memory for faces and associated perceptual versus social contexts.

The current study investigates cultural differences in memory for faces and their associated contexts between East Asian and Western cultures. Thirty-six Caucasian Canadian (age range = 17 - 30, M = 21.00, SD = 3.94) and 35 native Chinese (age range = 19 – 27, M = 22.06, SD = 2.33) young adults participated in the study. At encoding, participants subjectively rated face stimuli according to the cue words within social (“FRIENDLY” vs. “SUCCESS”) and perceptual (“HEIGHT” vs. “WEIGHT”) contexts. At test, participants performed a context recognition task to identify whether the test stimulus was a new face or an old face associated with a specific context cue word. Results did not show any cultural difference in memory. However, both cultures showed better recognition in the social context than in the perceptual context condition. In addition, the other race effect was observed in context memory for Canadian, but not Chinese participants.

Cultural differences in memory for faces and associated perceptual versus social contexts.

The current study investigates cultural differences in memory for faces and their associated contexts between East Asian and Western cultures. Thirty-six Caucasian Canadian (age range = 17 - 30, M = 21.00, SD = 3.94) and 35 native Chinese (age range = 19 – 27, M = 22.06, SD = 2.33) young adults participated in the study. At encoding, participants subjectively rated face stimuli according to the cue words within social (“FRIENDLY” vs. “SUCCESS”) and perceptual (“HEIGHT” vs. “WEIGHT”) contexts. At test, participants performed a context recognition task to identify whether the test stimulus was a new face or an old face associated with a specific context cue word. Results did not show any cultural difference in memory. However, both cultures showed better recognition in the social context than in the perceptual context condition. In addition, the other race effect was observed in context memory for Canadian, but not Chinese participants.

Altered activation and functional connectivity in individuals with social anhedonia when envisioning positive future episodes

Background Anticipatory pleasure deficits are closely correlated with negative symptoms in schizophrenia, and may be found in both clinical and subclinical populations along the psychosis continuum. Prospection, which is an important component of anticipatory pleasure, is impaired in individuals with social anhedonia (SocAnh). In this study, we examined the neural correlates of envisioning positive future events in individuals with SocAnh. Methods Forty-nine individuals with SocAnh and 33 matched controls were recruited to undergo functional MRI scanning, during which they were instructed to simulate positive or neutral future episodes according to cue words. Two stages of prospection were distinguished: construction and elaboration. Results Reduced activation at the caudate and the precuneus when prospecting positive (v. neutral) future events was observed in individuals with SocAnh. Furthermore, compared with controls, increased functional connectivity between the caudate and the inferior occipital gyrus during positive (v. neutral) prospection was found in individuals with SocAnh. Both groups exhibited a similar pattern of brain activation for the construction v. elaboration contrast, regardless of the emotional context. Conclusions Our results provide further evidence on the neural mechanism of anticipatory pleasure deficits in subclinical individuals with SocAnh and suggest that altered cortico-striatal circuit may play a role in anticipatory pleasure deficits in these individuals.

Learning new verbs with known cue words: The relative effects of noun and adverb cues

Research has shown that learning a known-and-unknown word combination leads to greater learning than learning an unknown word alone (Kasahara, 2010, 2011). These studies found that attaching a known adjective to an unknown noun can help learners remember the unknown noun. Kasahara (2015) found that a known verb can serve as an effective cue to remember an unknown noun in a known-and-unknown combination. To examine useful cues to learn unknown verbs, this study compared verb (unknown) + noun (known) combinations to verb (unknown) + adverb (known) combinations. Additionally, we explored how learners’ vocabulary size would affect the known-and-unknown two-word combination learning to deepen our understanding of the characteristics of students who benefit from combination learning. The participants in each group learned 18 two-word combinations consisting of the same unknown target verbs and different known cues (nouns or adverbs). The participants were provided with a five-minute learning phase and two immediate recall tests: a Single Word Test, to write down the L1 meanings of the targets, and a Combination Test, to write down the L1 meanings of the combinations. The same two tests were administered one week later. The results showed that known nouns were better cues for learning unknown verbs than known adverbs. It was also found that participants with a larger vocabulary size benefited more from two-word combination learning.

Interpretable Multi-Head Self-Attention Architecture for Sarcasm Detection in Social Media

With the online presence of more than half the world population, social media plays a very important role in the lives of individuals as well as businesses alike. Social media enables businesses to advertise their products, build brand value, and reach out to their customers. To leverage these social media platforms, it is important for businesses to process customer feedback in the form of posts and tweets. Sentiment analysis is the process of identifying the emotion, either positive, negative or neutral, associated with these social media texts. The presence of sarcasm in texts is the main hindrance in the performance of sentiment analysis. Sarcasm is a linguistic expression often used to communicate the opposite of what is said, usually something that is very unpleasant, with an intention to insult or ridicule. Inherent ambiguity in sarcastic expressions make sarcasm detection very difficult. In this work, we focus on detecting sarcasm in textual conversations from various social networking platforms and online media. To this end, we develop an interpretable deep learning model using multi-head self-attention and gated recurrent units. The multi-head self-attention module aids in identifying crucial sarcastic cue-words from the input, and the recurrent units learn long-range dependencies between these cue-words to better classify the input text. We show the effectiveness of our approach by achieving state-of-the-art results on multiple datasets from social networking platforms and online media. Models trained using our proposed approach are easily interpretable and enable identifying sarcastic cues in the input text which contribute to the final classification score. We visualize the learned attention weights on a few sample input texts to showcase the effectiveness and interpretability of our model.

Escaping from Revulsion - Disgust and Escape in Response to Body-Relevant Autobiographical Memories

The term Repulsive Body Image (RBI) refers to a schematic construct combining body-directed self-disgust and other negative body image features. As a self-schema, the RBI is assumed to bias information processing, including autobiographical memory retrieval. When specific memories about the own body are retrieved in a direct (automatic) fashion, intense self-disgust may arise. This may trigger attempts to escape from those memories which, in turn, might further strengthen the RBI. We asked 133 women with high (H-RBI; n = 63) and low (L-RBI; n = 70) levels of habitual body-directed self-disgust to recall autobiographical memories in response to 11 concrete body-related cue words in a minimal instructions Autobiographical Memory Test (AMT). Despite an overall low level of memory specificity, we found that RBI levels were associated with stronger disgust reactions and stronger tendencies to escape from body-related memories. In addition, increased disgust reactions to body-related memories accounted for the association between habitual levels of self-disgust and increased tendencies to escape from these memories. Thus, the findings indicated that women with body image concerns showed disgust-based escape from body-related memories. This disgust-based inclination to avert from body-related autobiographical memories might counteract the correction of an RBI, thereby contributing to the persistence of body image concerns and associated psychopathology.

Interpretable Multi-Head Self-Attention Architecture for Sarcasm Detection in Social Media

Sarcasm is a linguistic expression often used to communicate the opposite of what is said, usually something that is very unpleasant with an intention to insult or ridicule. Inherent ambiguity in sarcastic expressions, make sarcasm detection very difficult. In this work, we focus on detecting sarcasm in textual conversations from various social networking platforms and online media. To this end, we develop an interpretable deep learning model using multi-head self-attention and gated recurrent units. Multi-head self-attention module aids in identifying crucial sarcastic cue-words from the input, and the recurrent units learn long-range dependencies between these cue-words to better classify the input text. We show the effectiveness of our approach by achieving state-of-the-art results on multiple datasets from social networking platforms and online media. Models trained using our proposed approach are easily interpretable and enable identifying sarcastic cues in the input text which contribute to the final classification score. We visualize the learned attention weights on few sample input texts to showcase the effectiveness and interpretability of our model.