Relational Features Research Articles

Syntactic Information and Multiple Semantic Segments for Aspect-Based Sentiment Classification

Aspect-based sentiment classification (ASC) is a task to determine the sentiment polarities of specific aspects in a review. Syntactic information like dependency relation has been proven effective when extracting the sentiment features. On the other hand, multiple semantic segments in a review may influence the sentiment polarity. Thus, we propose a neural network based on dependency relation and structured attention (DRSAN) to fuse both dependency relation features and multiple semantic segments with different attention mechanisms. To verify the performance of DRSAN, we build a Chinese Mobile Phone Review (CMPR) dataset. To our knowledge, we are the first to explicitly integrate dependency relation and structured attention for the ASC task. The experimental results on SemEval 2014 Task 4, Twitter, CMPR, and several other cross-lingual and cross-domain datasets show that the proposed model outperforms all other benchmark models.

Knowledge-Enhanced Graph Attention Network for Fact Verification

Fact verification aims to evaluate the authenticity of a given claim based on the evidence sentences retrieved from Wikipedia articles. Existing works mainly leverage the natural language inference methods to model the semantic interaction of claim and evidence, or further employ the graph structure to capture the relation features between multiple evidences. However, previous methods have limited representation ability in encoding complicated units of claim and evidences, and thus cannot support sophisticated reasoning. In addition, a limited amount of supervisory signals lead to the graph encoder could not distinguish the distinctions of different graph structures and weaken the encoding ability. To address the above issues, we propose a Knowledge-Enhanced Graph Attention network (KEGA) for fact verification, which introduces a knowledge integration module to enhance the representation of claims and evidences by incorporating external knowledge. Moreover, KEGA leverages an auxiliary loss based on contrastive learning to fine-tune the graph attention encoder and learn the discriminative features for the evidence graph. Comprehensive experiments conducted on FEVER, a large-scale benchmark dataset for fact verification, demonstrate the superiority of our proposal in both the multi-evidences and single-evidence scenarios. In addition, our findings show that the background knowledge for words can effectively improve the model performance.

On the dual role of expressive speech acts: Relational work on signs announcing closures during the Covid-19 pandemic

This paper provides some new insights into the dual function of expressive speech acts discussed in pragmatic theory as either expressions of genuine emotions or conventionalised acts of courtesy. Drawing on the framework of interpersonal pragmatics, it analyses signs displayed on the doors of closed businesses in Athens and London during the first lockdown of the Covid-19 pandemic. These closure signs are characterised by a heavy use of relational features, including four expressive speech acts, namely greetings, apologies, thanks and wishes, which form the focus of the analysis. The relational work performed by these speech acts reflects the social changes brought about by the pandemic as well as the business owners’ attempts to retain their customers. The expressive speech acts featured on the signs are evaluated against the norms set out by the genre of closure signs. The comparative angle of the study, on the other hand, links the discussion to norms related to the display of emotions and to the use of conventionalised formulae in the two countries under study, thus revealing culture-specific perspectives on the dual function of expressive speech acts.

Learning hyperbolic attention-based embeddings for link prediction in knowledge graphs

Knowledge graph (KG) embedding methods aim to learn low-dimensional representations of entities and relations to predict new valid triples for KG completion. Most of the existing KG embedding models learn embeddings in Euclidean space, which cannot accurately capture hierarchical structures and complex properties of relations found in KGs. To this effect, a recent model MuRP, as a first hyperbolic method, learns the KG embeddings in hyperbolic space and outperforms existing Euclidean embedding models. However, MuRP treats the KG triples individually, and hence fails to capture the complex structural information inherent in the local vicinity of a node, leading to low-quality node embeddings. On the other hand, the recent hyperbolic graph neural network (HGNN) provides a way of learning high-quality hyperbolic node embeddings by capturing information from each node’s neighborhood. However, HGNN ignores the relation features and treats all neighboring nodes with equal importance. To this end, we propose RHGNN, which extends HGNN by including the relation features and performing a hyperbolic attention-based neighborhood aggregation. We then combine RHGNN with MuRP into a novel encoder–decoder hyperbolic embedding learning framework (which we call HyperGEL) for KG completion. RHGNN gathers information from the neighboring nodes to generate rich hyperbolic node embeddings, and MuRP uses these embeddings to predict new triples. Experimental results show the proposed framework’s effectiveness that consistently marks performance gains over several previous models on recent standard KG completion datasets.

Research on the mechanism of knowledge diffusion in the MOOC learning forum using ERGMs

Research on the mechanism of knowledge diffusion in the MOOC learning forum using ERGMs

Exchange and Communal Strengths’ Associations With Relational Characteristics and Forgiveness

Clark and Mills (2012) proposed that communal norms characterize “healthy marriages,” whereas exchange norms indicate “troubled” ones. Using the actor–partner interdependence model, we consider this proposal. Heterosexual married partners’ exchange and communal strength are examined with several relational features including trust, commitment, relational satisfaction, and control mutuality, as well as partner-specific tendencies toward forgiveness. The findings indicate that one’s communal strength is associated with positive relational features for both oneself and one’s spouse, as well as a greater propensity to forgive and lesser tendencies toward negative forgiveness (retaliation). Exchange strength was generally associated with negative forgiveness. However, the relationship between exchange strength and relational features is more complicated. Lesser communal strength was generally associated with lower levels of the relational characteristics. Yet, exchange and communal strength interacted, indicating exchange may play a protective function by buffering against the ill effects of a lack of communal strength for some relational characteristics.

What makes a façade beautiful?

The aesthetic of our built environment is perceived as an important aspect to consider for the design of human-centred cities, but a problem quickly arises in the presence of clashing conceptions of what we understand to be aesthetically pleasing. This paper adds to this discussion, by exploring architects’ aesthetic preferences in facade design, aiming to include design practitioners in a debate that so far has remained largely academic. Thus, the goal of the study was to identify relevant aspects, based on a series of semi-structured interviews with practitioners representing 34 architectural firms from The Netherlands, carried out from January to April 2020. It was possible to identify two major types of aspects, and subsequent sub-groups. Intrinsic aspects (compositional, plastic, detail design, and character) comprise aspects that are characteristic of a facade as an object, while extrinsic aspects (human, intellectual, and contextual connection) consist of relational features, determining the perceived beauty of a facade in terms of its connections with an outside agent. The main identified aspects in each one of the groups, the potential relations among them, and their relative relevance within the surveyed sample were part of the assessment, comparing the outcomes against previous results from the literature.

Suicidal ideation and mental disorder detection with attentive relation networks

Mental health is a critical issue in modern society, and mental disorders could sometimes turn to suicidal ideation without effective treatment. Early detection of mental disorders and suicidal ideation from social content provides a potential way for effective social intervention. However, classifying suicidal ideation and other mental disorders is challenging as they share similar patterns in language usage and sentimental polarity. This paper enhances text representation with lexicon-based sentiment scores and latent topics and proposes using relation networks to detect suicidal ideation and mental disorders with related risk indicators. The relation module is further equipped with the attention mechanism to prioritize more critical relational features. Through experiments on three real-world datasets, our model outperforms most of its counterparts.

Understanding ‘context’ in realist evaluation and synthesis

ABSTRACT Context is a key concept in developing realist causal explanations but its conceptualisation has received comparatively less attention. We conducted a review to explore how context is conceptualised within realist reviews and evaluations. We purposively selected 40 studies to examine: How is context defined? And how is context operationalised in the findings? We identified two key ‘narratives’ in the way context was conceptualised: 1) Context as observable features (space, place, people, things) that triggered or blocked the intervention; assuming that context operates at one moment in time and sets in motion a chain reaction of events. 2) Context as the relational and dynamic features that shaped the mechanisms through which the intervention works; assuming that context operates in a dynamic, emergent way over time at multiple different levels of the social system. These two context narratives have different implications for the design, goals and impact of realist reviews and evaluations.

Performing Identities and Performing Relationships: Community Music Therapy and Adolescent Mental Health

AbstractThis article explores the “Coffee House,” a community music therapy performance event held biannually at an adolescent mental health treatment facility in Southwestern Ontario, Canada. In this paper, I draw upon techniques and theory from narrative inquiry in order to investigate the experiences and perspectives of 7 adolescent clients and 11 staff members who participated in the event as performers and audience members. Data were collected through semi-structured interviews and analyzed through qualitative coding; the participants’ voices are presented here. Building upon a previous article, in which I attribute the Coffee House’s success to its participatory ethos, this article examines the impact of performing upon participants’ musical and personal identities as well as upon their relationships with others at the facility. The shifts and transformations that took place within youths’ identities were interdependent with the relational features of the performance context; expansions in youths’ self-identities were indelibly connected to staff members’ expanded perspectives on these youths, afforded through witnessing their performances. Participants’ narratives validate not only the ways in which identity and relationship intersect, but also the way in which musical performance’s impact upon identity and relationship is uniquely musical.

An adaptive viscous damping wall for seismic protection: Experimental study and performance-based design

In this paper, a novel adaptive viscous damping wall (A-VDW) system was developed to achieve a two-stage seismic mitigation. The A-VDW combines a viscous damping wall with lead dampers through working gaps. The A-VDW provides only damping but no extra stiffness for moderate or minor earthquakes, while for severe earthquakes, both stiffness and energy dissipation are provided. A numerical model which captures the displacement—force relation features at different working stages was proposed for the A-VDW; then, dynamic tests on a scaled A-VDW under different loading frequencies and amplitudes were conducted to verify the proposed physical configuration and numerical model. The two-stage seismic mitigation effect of the proposed A-VDW is evaluated in single-degree-of-freedom (SDOF) systems. Finally, a performance-based design method considering velocity exponent was proposed and also validated.

3DRM: Pair-wise relation module for 3D object detection

Context has proven to be one of the most important factors in object layout reasoning for 3D scene understanding. Existing deep contextual models either learn holistic features for context encoding or rely on pre-defined scene templates for context modeling. We argue that scene understanding benefits from object relation reasoning, which is capable of mitigating the ambiguity of 3D object detections and thus helps locate and classify the 3D objects more accurately and robustly. To achieve this, we propose a novel 3D relation module (3DRM) which reasons about object relations at pair-wise levels. The 3DRM predicts the semantic and spatial relationships between objects and extracts the object-wise relation features. We demonstrate the effects of 3DRM by plugging it into proposal-based and voting-based 3D object detection pipelines, respectively. Extensive evaluations show the effectiveness and generalization of 3DRM on 3D object detection. Our source code is available at https://github.com/lanlan96/3DRM.

Exploring relational and moral features in medical students

ObjectivesWe aimed to explore empathy, moral competencies, callous traits, and temperament in a sample of medical students. Furthermore, we aimed to investigate differences in our variables across the 1st and 5th years of medical education and possible correlations between them. MethodsThis was a cross-sectional study with 138 medical students. We resorted to self-reported instruments that were given at the end of classes: Barrett-Lennard Relational Inventory, Temperament Evaluation of Memphis, Pisa and San Diego Auto-questionnaire, Inventory of Callous-Unemotional Traits, and Moral Competence Test. For the statistical analysis, we resorted to descriptive and inferential statistics, using non-parametric tests when data didn't follow a normal distribution. ResultsWe found no statistical difference between empathy scores in 1st (N=104, Mean=41.42, SD=22.48) and 5th year students (N=34, Mean=37.35, SD=23.35), t(136)=0.908, p=0.366. Callous traits were negatively correlated with empathy (r(136)=-0.444, p=0.000) and no correlation between moral competences and empathy (r(96)=0.029, p=0.779) was observed. We found a negative correlation between empathy and cyclothymic, anxious and irritable temperaments (r(136)=-0.334, p=0.000, r(136)=-0.281, p=0.001, r(136)=-0.400, p=0.000). ConclusionsOur scores corroborate previous evidence that medical students are empathic, have good moral standards and low callous traits. We saw no differences in empathy scores between the two years and future studies could explore the particulars of medical curriculums impacting this variable. In our study, empathy was negatively correlated with callous traits and linked with specific temperaments. Considering these variables at admission to medical school as well as preserving and improving them in medical education might offer better standards of care.

A relational perspective on emotion regulation across adulthood

AbstractTheories of aging suggest that close social connections are key determinants of maintained emotional well‐being across adulthood. However, most emotion regulation (ER) research has taken an intrapsychic perspective, focusing primarily on individuals' attempts to manage their own emotions in isolation from others. In this article, we present a relational perspective on ER and suggest that consideration of social factors can provide novel insight into how ER operates and changes throughout the adult lifespan. We describe theoretical perspectives of aging and ER, review the mixed evidence regarding the characterization of age‐related shifts in ER, and highlight relational features of ER goals, strategy selection, and success of implementation. We conclude with a discussion of directions for future research examining ER throughout adulthood from a relational perspective, advocating for the comparison of intrapersonal and interpersonal ER in ecologically valid and longitudinal contexts.

AutoBOT: evolving neuro-symbolic representations for explainable low resource text classification

Learning from texts has been widely adopted throughout industry and science. While state-of-the-art neural language models have shown very promising results for text classification, they are expensive to (pre-)train, require large amounts of data and tuning of hundreds of millions or more parameters. This paper explores how automatically evolved text representations can serve as a basis for explainable, low-resource branch of models with competitive performance that are subject to automated hyperparameter tuning. We present autoBOT (automatic Bags-Of-Tokens), an autoML approach suitable for low resource learning scenarios, where both the hardware and the amount of data required for training are limited. The proposed approach consists of an evolutionary algorithm that jointly optimizes various sparse representations of a given text (including word, subword, POS tag, keyword-based, knowledge graph-based and relational features) and two types of document embeddings (non-sparse representations). The key idea of autoBOT is that, instead of evolving at the learner level, evolution is conducted at the representation level. The proposed method offers competitive classification performance on fourteen real-world classification tasks when compared against a competitive autoML approach that evolves ensemble models, as well as state-of-the-art neural language models such as BERT and RoBERTa. Moreover, the approach is explainable, as the importance of the parts of the input space is part of the final solution yielded by the proposed optimization procedure, offering potential for meta-transfer learning.

Automated segmentation for whole human eye OCT image using RM multistage Mask R-CNN

Accurate segmentation of structures in whole human eye optical coherence tomography (OCT) images can improve the accuracy of disease diagnosis and ophthalmic biological parameter measurement. However, due to the complex features of the cornea, lens, and retina in the whole human eye OCT image and external interference, it is difficult to precisely extract the object features, which restricts the segmentation accuracy of whole human eye OCT images. A relation module (RM) multistage Mask Region-based Convolutional Neural Network (R-CNN) method for whole human eye OCT image segmentation is established. Based on Mask R-CNN, modulated deformable convolution is employed to produce an irregular receptive field and improve the adaptability of the network to the whole human eye OCT image’s object deformation. In the feature map extraction, the RM is combined to learn the position relation feature map of the human eye structures, which is utilized to enhance feature maps. A multistage mask is constructed in the segmentation branch, and the error is corrected through iterations of the mask to improve the segmentation accuracy of the whole human eye OCT image. Combined with the above structures, the RM multistage Mask R-CNN method is constructed for whole human eye OCT image segmentation. The model is trained by whole human eye OCT images and is applied to realize highly accurate whole human eye OCT image segmentation. Comparison experiments with K-means, U-net, and Feature Pyramid Networks (FPN)-deformable-mask R-CNN are performed to verify the segmentation accuracy and robustness of the proposed method to complex deformation and external interference.

Covid‐19 and The Study of Professionals and Professional Work

Covid‐19 and The Study of Professionals and Professional Work

Pairwise acquisition prediction with SHAP value interpretation

Predicting future pairs of the acquirer and acquiree companies is important for acquisition or investment strategy. This prediction is a challenging problem due to the following requirements: to incorporate various non-financial factors and to address the lack of negative samples. Concerning the former, we proposed including a network feature that represented the importance of an acquirer and an acquiree in the investment and category networks, as well as a company relation feature associated with their similarity and closeness. Considering the latter requirement, as negative examples, we set the pairs of acquirers and acquirees with the features that were similar to those of positive examples. This allowed learning minor differences between the companies selected for acquisition and the candidate ones. We evaluated our proposed prediction model using 2000–2018 acquisition logs collected from CrunchBase. Based on the analysis of the high SHapley additive explanation (SHAP) value features, we found that the newly considered network and company relation features had high significance (10 out of 22 top key features). We also clarified how these novel features contributed to the prediction of acquisition occurrence by interpreting the SHAP value.

Frog-GNN: Multi-perspective aggregation based graph neural network for few-shot text classification

Few-shot text classification aims to learn a classifier from very few labeled text instances per class. Previous few-shot research works in NLP are mainly based on Prototypical Networks, which encode support set samples of each class to prototype representations and compute distance between query and each class prototype. In the prototype aggregation progress, much useful information of support set and discrepancy between samples from different class are ignored. In contrast, our model focuses on all query-support pairs without information loss. In this paper, we propose a multi-perspective aggregation based graph neural network (Frog-GNN) that observes through eyes (support and query instance) and speaks by mouth (pair) for few-shot text classification. We construct a graph by pre-trained pair representations and aggregate information from neighborhoods by instance-level representations for message-passing. The final relational features of pairs imply intra-class similarity and inter-class dissimilarity after iteratively interactions among instances. In addition, our Frog-GNN with meta-learning strategy can well generalize to unseen class. Experimental results demonstrate that the proposed GNN model outperforms existing few-shot approaches in both few-shot text classification and relation classification on three benchmark datasets.

Situational context influences the degree of hostile attributions made by individuals with schizophrenia or autism spectrum disorder.

Previous work has demonstrated that the tendency to make hostile attributions is not a stable trait but varies across different social situations. Therefore, we sought to investigate whether hostile attributions within clinical samples are better understood as a persistent characteristic or one that varies across contexts. The current analyses investigated patterns of attributions among people diagnosed with schizophrenia (SCZ, n=271) or autism spectrum disorder (ASD, n=100) and non-clinical control participants (NCC, n=233) in an existing data set. Results showed that specific relational features in vignettes portraying different social encounters influence the way people make attributions and that variability across contexts is present in both non-clinical and clinical populations. Like non-clinical participants, participants diagnosed with ASD ascribed the greatest hostility to a scene involving an authority figure. In contrast, SCZ participants reported the greatest hostility in response to a scene involving a friend. These findings suggest that salient environmental factors should be considered when assessing social cognitive skills and biases. Hostile attributions should be perceived as situational constructs rather than stable and persistent characteristics. Hostile attributions were most prevalent among persons diagnosed with schizophrenia; however, on average, all participants showed greater hostility for situations involving an authority figure, an acquaintance, or a friend relative to those involving a co-worker or stranger. Psychotherapists and clinicians working with people diagnosed with schizophrenia or autism spectrum disorder could work on identifying situation triggers, which may prompt hostile attributions. Psycho-educational and psychotherapeutic interventions can be altered based on individual triggers of hostile attributions, and attempts can be made to lessen these attributions. Paranoia appears to be linked to hostile attributions regardless of the specific clinical diagnosis and should be considered in the therapeutic process.