Context-dependent Features Research Articles

On Trust Recommendations in the Social Internet of Things – A Survey

The novel paradigm Social Internet of Things (SIoT) improves the network navigability, identifies suitable service providers, and addresses scalability concerns. Ensuring trustworthy collaborations among devices is a key aspect in SIoT and can be realized through trust recommendations. However, the outcome of trust recommendations depends on multiple factors related to the context-dependent nature of SIoT and practical constraints brought by the devices and networks embedded in the SIoT. While the existing literature has proposed numerous trust recommendation models to assess the trustworthiness of devices in various scenarios, researchers have not sufficiently examined the required features for trust recommendations in the SIoT. Consequently, trust recommendation models may inaccurately assess the true risk of device interactions. In this literature survey, we investigate the context-dependent features and recommendation methods used for the SIoT using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology. We propose a novel taxonomy to categorize trust recommendation models according to their input features and design. Our findings reveal limited attention is given to the context-dependent features, constraints of the information environment, and limited inference capabilities that impede more precise trust recommendations. Finally, we present the research gaps and outline future directions to enable trustworthy inter-domain operations within the SIoT.

Warming-induced flowering and fruiting in the seagrass Posidonia oceanica and uncertainties due to context-dependent features

This study explores the impacts of global warming on the reproductive efforts of Posidonia oceanica, a key seagrass species endemic to the Mediterranean Sea. By analyzing data from five areas in Sardinia in February 2023, we have investigated the relationship between thermal descriptors and P. oceanica flowering and fruiting variables. In fact, whether climate change has been affected the distribution of seagrasses by changing their reproductive efforts is still controversial and our findings contribute to this ongoing debate, highlighting a noticeable increase in reproductive events potentially affected by rising temperature. Overall, mean summer sea surface temperature (SST) was positively associated to inflorescence density, presence of living fruits and fruit size; living fruits and fruit size were also influenced positively by the maximum SST within 3 months before sampling (November–February). Both flowering and fruiting were negatively affected by the maximum SST within 1 month before inflorescence observation and mean winter SST, respectively, suggesting that P. oceanica reproductive effort depends on the duration and the temporal context in which the anomaly occurs. However, the high variability between meadows of the same area highlights the need of future studies to focus on the uncertainties related to local factors.

A study on classifying Stack Overflow questions based on difficulty by utilizing contextual features

Technical question-answering sites like Stack Overflow are gaining enormous attention from practitioners of specialized fields looking to exchange their programming knowledge. They ask questions on different topics with varying degrees of complexity and difficulty. All practitioners do not have the same level of expertise on those topics to respond to such questions. However, the current approach used by Stack Overflow mostly filters questions based on topics alone and does not take difficulty into account. For this reason, a large percentage of questions fail to attract the attention of appropriate users, resulting in questions having no answer or a significant delay in response time. To address these limitations, we incorporate three models, TF-IDF, LDA, and Doc2Vec, to extract semantic and context-dependent features that can measure the difficulty of questions. Each of these models is paired with different classifiers along with other features to classify the questions based on difficulty. Extensive experiments on three different datasets exhibit the effectiveness of our models, and Doc2Vec outperforms the other models. We also identified that the contextual features are correlated with question difficulty, and one subset of features outperforms others. The proposed approach can be beneficial for building an automatic tagger based on question difficulty.

Chromatin structure and context-dependent sequence features control prime editing efficiency.

Prime editing (PE) is a highly versatile CRISPR-Cas9 genome editing technique. The current constructs, however, have variable efficiency and may require laborious experimental optimization. This study presents statistical models for learning the salient epigenomic and sequence features of target sites modulating the editing efficiency and provides guidelines for designing optimal PEs. We found that both regional constitutive heterochromatin and local nucleosome occlusion of target sites impede editing, while position-specific G/C nucleotides in the primer-binding site (PBS) and reverse transcription (RT) template regions of PE guide RNA (pegRNA) yield high editing efficiency, especially for short PBS designs. The presence of G/C nucleotides was most critical immediately 5' to the protospacer adjacent motif (PAM) site for all designs. The effects of different last templated nucleotides were quantified and observed to depend on the length of both PBS and RT templates. Our models found AGG to be the preferred PAM and detected a guanine nucleotide four bases downstream of the PAM to facilitate editing, suggesting a hitherto-unrecognized interaction with Cas9. A neural network interpretation method based on nonextensive statistical mechanics further revealed multi-nucleotide preferences, indicating dependency among several bases across pegRNA. Our work clarifies previous conflicting observations and uncovers context-dependent features important for optimizing PE designs.

DIPG-50. INSIGHTS INTO TUMORIGENESIS OF DIFFUSE INTRINSIC PONTINE GLIOMA-ONCOHISTONE AND SIGNALING

Abstract Diffuse intrinsic pontine glioma (DIPG) is a rare and incurable pediatric brain cancer with survival of less than one year. Understanding the tumorigenesis mechanisms of DIPG and identifying potential therapeutic strategy are major research foci in the DIPG field. The most lethal subtype of diffuse intrinsic pontine glioma (DIPG) is the H3K27M subtype. Although ACVR1 mutations have been implicated in the pathogenesis of this presently incurable disease, the impacts of BMP signaling on more than 60% of H3K27M DIPG carrying ACVR1 wild-type remain unknown. Here, we show that BMP ligands exert potent tumor suppressive effects against H3.3 K27M and ACVR1 WT DIPG in a SMAD-dependent manner. Specifically, clinical data revealed that many DIPG tumors have exploited CHRDL1’s capacity to hijack BMP ligands. We discovered activation of BMP signaling promotes the exit of DIPG tumor cells from “prolonged-stem-cell-like” state to differentiation by epigenetically regulating CXXC5, which acts as a tumor suppressor and positive regulator of BMP signaling. Beyond showing how BMP signaling impacts DIPG, our study also identified potent anti-tumor efficacy of Dacinostat for DIPG. Thus, our study delineates context-dependent features of BMP signaling pathway in DIPG subtype.

A Novel Deep Learning Method for Automatic Recognition of Coseismic Landslides

Massive earthquakes generally trigger thousands of coseismic landslides. The automatic recognition of these numerous landslides has provided crucial support for post-earthquake emergency rescue, landslide risk mitigation, and city reconstruction. The automatic recognition of coseismic landslides has always been a difficult problem due to the relatively small size of a landslide and various complicated environmental backgrounds. This work proposes a novel semantic segmentation network, EGCN, to improve the landslide identification accuracy. EGCN conducts coseismic landslide recognition by a recognition index set as the input data, CGBlock as the basic module, and U-Net as the baseline. The CGBlock module can extract the relatively stable global context-dependent features (global context features) and the unstable local features by the GNN Branch and CNN Branch (GNN Branch contains the proposed EISGNN) and integrates them via adaptive weights. This method has four advantages. (1) The recognition indices are established according to the causal mechanism of coseismic landslides. The rationality of the indices guarantees the accuracy of landslide recognition. (2) The module of EISGNN is suggested based on the entropy importance coefficient and GATv2. Owing to the feature aggregation among nodes with high entropy importance, global and useful context dependency can be synthesized and the false alarm of landslide recognition can be reduced. (3) CGBlock automatically integrates context features and local spatial features, and has strong adaptability for the recognition of coseismic landslides located in different environments. (4) Owing to CGBlock being the basic module and U-Net being the baseline, EGCN can integrate the context features and local spatial characteristics at both high and low levels. Thus, the accuracy of landslide recognition can be improved. The meizoseismal region of the Ms 7.0 Jiuzhaigou earthquake is selected as an example to conduct coseismic landslide recognition. The values of the precision indices of Overall Accuracy, mIoU, Kappa, F1-score, Precision, and Recall reached 0.99854, 0.99709, 0.97321, 0.97396, 0.97344, and 0.97422, respectively. The proposed method outperforms the current major deep learning methods.

Real-Time 3D Single Object Tracking With Transformer

LiDAR-based 3D single object tracking is a challenging issue in robotics and autonomous driving. Currently, existing approaches usually suffer from the problem that objects at long distance often have very sparse or partially-occluded point clouds, which makes the features extracted by the model ambiguous. Ambiguous features will make it hard to locate the target object and finally lead to bad tracking results. To solve this problem, we utilize the powerful Transformer architecture and propose a Point-Track-Transformer (PTT) module for point cloud-based 3D single object tracking task. Specifically, PTT module generates fine-tuned attention features by computing attention weights, which guides the tracker focusing on the important features of the target and improves the tracking ability in complex scenarios. To evaluate our PTT module, we embed PTT into the dominant method and construct a novel 3D SOT tracker named PTTNet. In PTT-Net, we embed PTT into the voting stage and proposal generation stage, respectively. PTT module in the voting stage could model the interactions among point patches, which learns context-dependent features. Meanwhile, PTT module in the proposal generation stage could capture the contextual information between object and background. We evaluate our PTT-Net on KITTI and NuScenes datasets. Experimental results demonstrate the effectiveness of PTT module and the superiority of PTT-Net, which surpasses the baseline by a noticeable margin, ~10% in the Car category. Meanwhile, our method also has a significant performance improvement in sparse scenarios. In general, the combination of transformer and tracking pipeline enables our PTT-Net to achieve state-of-the-art performance on both two datasets. Additionally, PTT-Net could run in real-time at 40FPS on NVIDIA 1080Ti GPU. Our code is open-sourced for the research community at https://github.com/shanjiayao/PTT.

Paralanguage and ad hoc concepts

Abstract Ad hoc concept construction is regarded as a case of free pragmatic enrichment, so it is presented as a non-linguistically mandated process that is automatically accomplished during mutual parallel adjustment. Recent research suggests that this lexical pragmatic process may be marked and steered by various linguistic elements. These include evaluative morphemes, lexical and phrasal items adjacent to content words, and stylistic resources like repetition or rewording. This paper argues that paralanguage may fulfil a similar enacting function and finetune the conceptual representations arising from content words on the grounds of idiosyncratic, context-dependent features or shades, as well as propositional and non-propositional information about the speaker’s psychological states. However, the paper restricts this function to expressive interjections, prosodic inputs like pitch, contrastive stress and pace or tempo, and gestural inputs such as language-like gestures, pantomimes and emblems. Conative interjections, intonation and proper gesticulation would be excluded from contributing to lexical pragmatic processes.

Vocal behaviour reveals asymmetries in neighbour relationships in a semi-colonial raptor, the Eurasian Scops Owl Otus scops

Abstract Territorial animals often reduce aggression towards familiar neighbours compared to unfamiliar conspecifics. However, variation in the response to different neighbours is less known. In this work, I examined the territorial behaviour of male scops owls during countersinging interactions with two familiar neighbours and I asked whether vocal behaviour of the focal male reflected dear-enemy relationships. Analysis revealed that the focal male’s vocal frequency was associated with (1) the degree of instability of the territory boundary shared with a neighbour and (2) the motivation to persist in the dyadic interaction with that neighbour. Patterns of movement directed to specific individuals suggest that scops owls do discriminate between neighbours. A case of partial territory takeover was observed that was accompanied by temporal changes in vocal frequency in one of the opponents, confirming that vocal frequency is a flexible, context-dependent feature of the relationship of neighbouring scops owls.

A Deep Learning-Based Assisted Teaching System for Oral English

The progress of global economic integration has forced English learners to have an urgent need to improve their oral English. College students’ oral English ability is currently the worst of the four abilities of listening, speaking, reading, and writing. The main reasons are internal and external. The internal reason is that the pronunciation characteristics of Chinese students are different from those of English. The external cause is that the practice environment and tools of oral English are not ideal, which affects the improvement of learners’ oral English. This study proposes using a deep learning algorithm (DLA) English in the evaluation of oral English quality to improve learners’ oral English level. The quality of oral English can be comprehensively evaluated in terms of pitch, speed of sound, and rhythm. The standard of pronunciation is the foundation of oral English and is the most critical factor. In many DLAs, the input unit of DNN at a certain moment and its upper and lower moment input units have no relationship and are independent of each other, and the timing dependencies of adjacent units are not fully considered. The results are generally not very good on speech recognition tasks. This study proposes a time-delay neural network (TDNN) and a long short-term memory (LSTM) network to calculate the posterior probability of the model state to model context-dependent features in order to solve this problem. The fusion model TDNN-LSTM is applied in the English spoken pronunciation recognition task. To compare the accuracy of oral English pronunciation, several classic DLAs are introduced. The experimental results show that the method described in this study has a number of advantages. Although the performance improvement of this method in terms of recognition accuracy is not large, a certain degree of improvement is also very important for the oral English teaching assistant system.

Context-dependent tumor-suppressive BMP signaling in diffuse intrinsic pontine glioma regulates stemness through epigenetic regulation of CXXC5.

The most lethal subtype of diffuse intrinsic pontine glioma (DIPG) is H3K27M. Although ACVR1 mutations have been implicated in the pathogenesis of this currently incurable disease, the impacts of bone morphogenetic protein (BMP) signaling on more than 60% of H3K27M DIPG carrying ACVR1 wild-type remain unknown. Here we show that BMP ligands exert potent tumor-suppressive effects against H3.3K27M and ACVR1 WT DIPG in a SMAD-dependent manner. Specifically, clinical data revealed that many DIPG tumors have exploited the capacity of CHRDL1 to hijack BMP ligands. We discovered that activation of BMP signaling promotes the exit of DIPG tumor cells from 'prolonged stem-cell-like' state to differentiation by epigenetically regulating CXXC5, which acts as a tumor suppressor and positive regulator of BMP signaling. Beyond showing how BMP signaling impacts DIPG, our study also identified the potent antitumor efficacy of Dacinostat for DIPG. Thus, our study delineates context-dependent features of the BMP signaling pathway in a DIPG subtype.

SOX2 transcription factor binding and function.

The transcription factor SOX2 is a vital regulator of stem cell activity in various developing and adult tissues. Mounting evidence has demonstrated the importance of SOX2 in regulating the induction and maintenance of stemness as well as in controlling cell proliferation, lineage decisions and differentiation. Recent studies have revealed that the ability of SOX2 to regulate these stem cell features involves its function as a pioneer factor, with the capacity to target nucleosomal DNA, modulate chromatin accessibility and prepare silent genes for subsequent activation. Moreover, although SOX2 binds to similar DNA motifs in different stem cells, its multifaceted and cell type-specific functions are reliant on context-dependent features. These cell type-specific properties include variations in partner factor availability and SOX2 protein expression levels. In this Primer, we discuss recent findings that have increased our understanding of how SOX2 executes its versatile functions as a master regulator of stem cell activities.

How Level is the 'Cognitive Playing Field'?

In Philosophy of Psychedelics, Letheby provides a convincing basis for the idea that psychedelics primarily derive their therapeutic potential through mediating favourable changes to self-related belief systems. In this commentary, we take a closer look at the role that contextual factors (‘set’ and ‘setting’) play in Letheby’s two-factor account of psychedelic therapy. While Letheby acknowledges that psychedelic effects are highly context-dependent, the exact role that context plays in self-modelling during the acute experience is not entirely clear. We argue that context plays an essential role in shaping the ‘discovery’ of alternate self-conceptions during the acute experience. Specifically, users are more likely to experience alterations in self-conception that are consistent with context-dependent features, such as one’s prior mindset (‘set’) and aspects of the external environment (‘setting’). This is consistent with a REBUS model of psychedelic effects, where a corollary effect of relaxing high-level prior beliefs is that the system becomes increasingly sensitised to bottom-up information (‘prediction errors’) from lower-level intrinsic sources and the sensory periphery. Furthermore, it may explain why a positive well-intentioned state of mind and a supportive therapeutic environment are more likely to result in positive acute experiences and long-term improvements in mental wellbeing. We see this position as being largely compatible with Letheby’s proposal, but may shift some aspects of its emphasis and framing. In particular, it may conflict with Letheby’s notion that psychedelics lead to a ‘level cognitive playing field’, where self-related hypotheses are thought to be assigned more or less equal probability.

From a Dynamic Image Annotation Process within the Humanities to a Canonical Workflow

Abstract One idea of the Canonical Workflow Framework for Research (CWFR) is to improve the reusability and automation in research. In this paper, we aim to deliver a concrete view on the application of CWFRs to a use case of the arts and humanities to enrich further discussions on the practical realization of canonical workflows and the benefits that come with it. This use case involves context dependent data transformation and feature extraction, ingests into multiple repositories as well as a “human-in-the-loop” workflow step, which introduces a certain complexity into the mapping to a canonical workflow.

Is free enrichment always free? Revisiting ad hoc-concept construction

In relevance-theoretic pragmatics, lexical adjustment, or ad hoc-concept construction, is considered a case of free pragmatic enrichment. It is hence included within the group of non-linguistically mandated processes that are automatically carried out during comprehension. Its output is an inferentially-specified representation of the conceptual content encoded, or activated, by open-class words like nouns, adjectives, verbs or adverbs. Ad hoc-concept construction is a necessary step to obtain a fully propositional, truth-evaluable form that captures in a more precise manner the portion of meaning that the speaker communicates explicitly. However, this paper argues that this lexical pragmatic process needs not always be described as completely exempt from linguistic mandate: it may sometimes be enacted and steered by a series of linguistic elements. These fall within two categories: (i) the evaluative morphemes that some lexical items may receive in some languages, and (ii) lexical and phrasal items adjacent to content words. These elements would finetune the representations arising from content words as regards more specific, contextdependent features or shades. Moreover, they could add to such representations information about the speaker's attitude towards, or stance about, what the speakers refer to. Therefore, that information needs not always be derived as implicatures.

CircRNA-binding protein site prediction based on multi-view deep learning, subspace learning and multi-view classifier.

Circular RNAs (circRNAs) generally bind to RNA-binding proteins (RBPs) to play an important role in the regulation of autoimmune diseases. Thus, it is crucial to study the binding sites of RBPs on circRNAs. Although many methods, including traditional machine learning and deep learning, have been developed to predict the interactions between RNAs and RBPs, and most of them are focused on linear RNAs. At present, few studies have been done on the binding relationships between circRNAs and RBPs. Thus, in-depth research is urgently needed. In the existing circRNA-RBP binding site prediction methods, circRNA sequences are the main research subjects, but the relevant characteristics of circRNAs have not been fully exploited, such as the structure and composition information of circRNA sequences. Some methods have extracted different views to construct recognition models, but how to efficiently use the multi-view data to construct recognition models is still not well studied. Considering the above problems, this paper proposes a multi-view classification method called DMSK based on multi-view deep learning, subspace learning and multi-view classifier for the identification of circRNA-RBP interaction sites. In the DMSK method, first, we converted circRNA sequences into pseudo-amino acid sequences and pseudo-dipeptide components for extracting high-dimensional sequence features and component features of circRNAs, respectively. Then, the structure prediction method RNAfold was used to predict the secondary structure of the RNA sequences, and the sequence embedding model was used to extract the context-dependent features. Next, we fed the above four views' raw features to a hybrid network, which is composed of a convolutional neural network and a long short-term memory network, to obtain the deep features of circRNAs. Furthermore, we used view-weighted generalized canonical correlation analysis to extract four views' common features by subspace learning. Finally, the learned subspace common features and multi-view deep features were fed to train the downstream multi-view TSK fuzzy system to construct a fuzzy rule and fuzzy inference-based multi-view classifier. The trained classifier was used to predict the specific positions of the RBP binding sites on the circRNAs. The experiments show that the prediction performance of the proposed method DMSK has been improved compared with the existing methods. The code and dataset of this study are available at https://github.com/Rebecca3150/DMSK.

Abstract P450: Single-cell Transcriptomic Profiling Of The Type-2 Diabetic Mouse Heart

Background: Diabetes is associated with a significantly elevated risk of heart failure. However, the precise cellular and molecular protagonists underpinning the development of heart failure in diabetes remains unclear. Moreover, very little is known, of how disparate non-myocyte populations of the heart contribute to diabetic cardiomyopathy. Methodology: To address this gap in knowledge, we conducted single-cell transcriptomic analysis of non-myocytes from heart ventricles of spontaneous type-2 diabetic ( db/db ) male mice. Findings were corroborated by flow cytometry, histology and computational analysis of publically available bulk RNA sequencing datasets from alternative diabetes models. Results: Single-cell transcriptomic analysis of db/db mouse hearts revealed a concerted diabetes-induced cellular response driving cardiac pathological remodelling. We identified diabetes-induced up-regulation of pathways contributing to known features of diabetic cardiomyopathy such as dysregulation of vascular homeostasis and lipid metabolism, as well as augmented inflammation, in cell specific contexts. We also identified unexpected characteristics in the diabetic heart, including impaired protein folding and cellular trafficking within lymphatic vessels. Using flow cytometry and histology, increase in inflammatory cells, such as Ly6C hi monocytes, shifts in macrophage phenotype, and increased abundances of fibroblasts and endocardial cells were confirmed. Finally, integration of single-cell transcriptomic data with publically available bulk-RNA sequencing datasets from alternative diabetes models revealed shared hallmarks of diabetic heart and disease context-dependent features. Conclusions: Together, this work offers a new perspective for understanding the cellular and molecular mediators of diabetes-induced cardiac pathology. Targeting these mediators may offer new therapeutic avenues to address the cardiac complications associated with diabetes.

Deciphering Plant Chromatin Regulation via CRISPR/dCas9-Based Epigenome Engineering.

CRISPR-based epigenome editing uses dCas9 as a platform to recruit transcription or chromatin regulators at chosen loci. Despite recent and ongoing advances, the full potential of these approaches to studying chromatin functions in vivo remains challenging to exploit. In this review we discuss how recent progress in plants and animals provides new routes to investigate the function of chromatin regulators and address the complexity of associated regulations that are often interconnected. While efficient transcriptional engineering methodologies have been developed and can be used as tools to alter the chromatin state of a locus, examples of direct manipulation of chromatin regulators remain scarce in plants. These reports also reveal pitfalls and limitations of epigenome engineering approaches that are nevertheless informative as they are often associated with locus- and context-dependent features, which include DNA accessibility, initial chromatin and transcriptional state or cellular dynamics. Strategies implemented in different organisms to overcome and even take advantage of these limitations are highlighted, which will further improve our ability to establish the causality and hierarchy of chromatin dynamics on genome regulation.

A feature-oriented vehicle trajectory data processing scheme for data mining: A case study for Statewide truck parking behaviors

We propose a workflow for trajectory data mining jointly using well-tested (as opposed to ad hoc) machine learning algorithms and unstructured local knowledge of experts and decision-makers, a common requirement in public agencies and consulting businesses. The key step of the workflow is to condense vehicle trajectory data into an analytics base table (ABT) using a set of features so that general-purpose data mining algorithms can be utilized. The case study extracts context-dependent features from high-frequency truck trajectory data from the State of Texas for analyzing patterns of truck parking in the Statewide highway system and for deriving implications for truck parking regulations and investment decisions. The results show that the approach is suitable for time-efficient implementation and provides valuable inputs for applications related to truck parking studies. This paper does not focus on the deeper understanding of the data in the case study; instead, it focuses on demonstrating how the proposed feature-oriented workflow eases the handling of high-volume trajectory data and improves the trackability of the decision process where data mining algorithms and human expertise interact significantly.

Riboexp: an interpretable reinforcement learning framework for ribosome density modeling.

Translation elongation is a crucial phase during protein biosynthesis. In this study, we develop a novel deep reinforcement learning-based framework, named Riboexp, to model the determinants of the uneven distribution of ribosomes on mRNA transcripts during translation elongation. In particular, our model employs a policy network to perform a context-dependent feature selection in the setting of ribosome density prediction. Our extensive tests demonstrated that Riboexp can significantly outperform the state-of-the-art methods in predicting ribosome density by up to 5.9% in terms of per-gene Pearson correlation coefficient on the datasets from three species. In addition, Riboexp can indicate more informative sequence features for the prediction task than other commonly used attribution methods in deep learning. In-depth analyses also revealed the meaningful biological insights generated by the Riboexp framework. Moreover, the application of Riboexp in codon optimization resulted in an increase of protein production by around 31% over the previous state-of-the-art method that models ribosome density. These results have established Riboexp as a powerful and useful computational tool in the studies of translation dynamics and protein synthesis. Availability: The data and code of this study are available on GitHub: https://github.com/Liuxg16/Riboexp. Contact:zengjy321@tsinghua.edu.cn; songsen@tsinghua.edu.cn.