Rich Knowledge Research Articles

Methods for evaluating unsupervised vector representations of genomic regions.

Representation learning models have become a mainstay of modern genomics. These models are trained to yield vector representations, or embeddings, of various biological entities, such as cells, genes, individuals, or genomic regions. Recent applications of unsupervised embedding approaches have been shown to learn relationships among genomic regions that define functional elements in a genome. Unsupervised representation learning of genomic regions is free of the supervision from curated metadata and can condense rich biological knowledge from publicly available data to region embeddings. However, there exists no method for evaluating the quality of these embeddings in the absence of metadata, making it difficult to assess the reliability of analyses based on the embeddings, and to tune model training to yield optimal results. To bridge this gap, we propose four evaluation metrics: the cluster tendency score (CTS), the reconstruction score (RCS), the genome distance scaling score (GDSS), and the neighborhood preserving score (NPS). The CTS and RCS statistically quantify how well region embeddings can be clustered and how well the embeddings preserve information in training data. The GDSS and NPS exploit the biological tendency of regions close in genomic space to have similar biological functions; they measure how much such information is captured by individual region embeddings in a set. We demonstrate the utility of these statistical and biological scores for evaluating unsupervised genomic region embeddings and provide guidelines for learning reliable embeddings.

LIANA+ provides an all-in-one framework for cell–cell communication inference

The growing availability of single-cell and spatially resolved transcriptomics has led to the development of many approaches to infer cell–cell communication, each capturing only a partial view of the complex landscape of intercellular signalling. Here we present LIANA+, a scalable framework built around a rich knowledge base to decode coordinated inter- and intracellular signalling events from single- and multi-condition datasets in both single-cell and spatially resolved data. By extending and unifying established methodologies, LIANA+ provides a comprehensive set of synergistic components to study cell–cell communication via diverse molecular mediators, including those measured in multi-omics data. LIANA+ is accessible at https://github.com/saezlab/liana-py with extensive vignettes (https://liana-py.readthedocs.io/) and provides an all-in-one solution to intercellular communication inference.

Enhancing source code classification effectiveness via prompt learning incorporating knowledge features

Researchers have investigated the potential of leveraging pre-trained language models, such as CodeBERT, to enhance source code-related tasks. Previous methodologies have relied on CodeBERT’s ‘[CLS]’ token as the embedding representation of input sequences for task performance, necessitating additional neural network layers to enhance feature representation, which in turn increases computational expenses. These approaches have also failed to fully leverage the comprehensive knowledge inherent within the source code and its associated text, potentially limiting classification efficacy. We propose CodeClassPrompt, a text classification technique that harnesses prompt learning to extract rich knowledge associated with input sequences from pre-trained models, thereby eliminating the need for additional layers and lowering computational costs. By applying an attention mechanism, we synthesize multi-layered knowledge into task-specific features, enhancing classification accuracy. Our comprehensive experimentation across four distinct source code-related tasks reveals that CodeClassPrompt achieves competitive performance while significantly reducing computational overhead.

From text to mask: Localizing entities using the attention of text-to-image diffusion models

Diffusion models have revolted the field of text-to-image generation recently. The unique way of fusing text and image information contributes to their remarkable capability of generating highly text-related images. From another perspective, these generative models imply clues about the precise correlation between words and pixels. This work proposes a simple but effective method to utilize the attention mechanism in the denoising network of text-to-image diffusion models. Without additional training time nor inference-time optimization, the semantic grounding of phrases can be attained directly. We evaluate our method on Pascal VOC 2012 and Microsoft COCO 2014 under weakly-supervised semantic segmentation setting and our method achieves superior performance to prior methods. In addition, the acquired word-pixel correlation is generalizable for the learned text embedding of customized generation methods, requiring only a few modifications. To validate our discovery, we introduce a new practical task called “personalized referring image segmentation” with a new dataset. Experiments in various situations demonstrate the advantages of our method compared to strong baselines on this task. In summary, our work reveals a novel way to extract the rich multi-modal knowledge hidden in diffusion models for segmentation.

InteractOA: Showcasing the representation of knowledge from scientific literature in Wikidata

Knowledge generated during the scientific process is still mostly stored in the form of scholarly articles. This lack of machine-readability hampers efforts to find, query, and reuse such findings efficiently and contributes to today’s information overload. While attempts have been made to semantify journal articles, widespread adoption of such approaches is still a long way off. One way to demonstrate the usefulness of such approaches to the scientific community is by showcasing the use of freely available, open-access knowledge graphs such as Wikidata as sustainable storage and representation solutions. Here we present an example from the life sciences in which knowledge items from scholarly literature are represented in Wikidata, linked to their exact position in open-access articles. In this way, they become part of a rich knowledge graph while maintaining clear ties to their origins. As example entities, we chose small regulatory RNAs (sRNAs) that play an important role in bacterial and archaeal gene regulation. These post-transcriptional regulators can influence the activities of multiple genes in various manners, forming complex interaction networks. We stored the information on sRNA molecule interaction taken from open-access articles in Wikidata and built an intuitive web interface called InteractOA, which makes it easy to visualize, edit, and query information. The tool also links information on small RNAs to their reference articles from PubMed Central on the statement level. InteractOA encourages researchers to contribute, save, and curate their own similar findings. InteractOA is hosted at https://interactoa.zbmed.de and its code is available under a permissive open source licence. In principle, the approach presented here can be applied to any other field of research.

Outlier detection based on multisource information fusion in incomplete mixed data

Multisource incomplete mixed data fusion (MsIMDF) plays a crucial role in outlier detection by utilizing complementary, informative, interpretative, and less noisy single-source data to identify unexpected errors or behaviors. However, existing multisource data fusion approaches only consider homogeneous data and single uncertain information,overlooking the mixed heterogeneous data and diverse uncertainty information. This limitation can negatively impact the performance of outlier detection. To address this issue, we propose MsIMDF-USF, a novel two-stage model that fully leverages the rich multisource knowledge and uncertainty information in incomplete mixed data. During the information fusion stage, the MsIMDF model combines multisource data into new single-source data using the minimum uncertainty strategy based on rough and fuzzy information. Subsequently, in the outlier detection stage, we reconstruct a neighborhood information network under a united-similar-fuzzy (USF) relationship using the new fused data. This reconstruction aims to strengthen the connections between similar objects while weakening relationships among dissimilar ones by considering single-attribute and multi-attribute information. Outlier scores are obtained based on the stationary distribution of the reconstructed networks using a Markov random walk. Experimental results on 16 real datasets demonstrate that the MsIMDF-USF model effectively extracts higher-quality data, exhibiting high applicability and robustness in outlier detection tasks.

Enhancing the vision–language foundation model with key semantic knowledge-emphasized report refinement

Recently, vision–language representation learning has made remarkable advancements in building up medical foundation models, holding immense potential for transforming the landscape of clinical research and medical care. The underlying hypothesis is that the rich knowledge embedded in radiology reports can effectively assist and guide the learning process, reducing the need for additional labels. However, these reports tend to be complex and sometimes even consist of redundant descriptions that make the representation learning too challenging to capture the key semantic information. This paper develops a novel iterative vision–language representation learning framework by proposing a key semantic knowledge-emphasized report refinement method. Particularly, raw radiology reports are refined to highlight the key information according to a constructed clinical dictionary and two model-optimized knowledge-enhancement metrics. The iterative framework is designed to progressively learn, starting from gaining a general understanding of the patient’s condition based on raw reports and gradually refines and extracts critical information essential to the fine-grained analysis tasks. The effectiveness of the proposed framework is validated on various downstream medical image analysis tasks, including disease classification, region-of-interest segmentation, and phrase grounding. Our framework surpasses seven state-of-the-art methods in both fine-tuning and zero-shot settings, demonstrating its encouraging potential for different clinical applications.

Source domain prior-assisted segment anything model for single domain generalization in medical image segmentation

Deep learning-based medical image segmentation models often suffer from performance degradation across domains due to domain discrepancies arising from data collected by various healthcare centers. Recent advancements, particularly the Segment Anything Model (SAM), have shown promising generalization abilities in this field, inspiring the development of several SAM-based approaches to address domain discrepancy issue. Nevertheless, these methods seldom exploit the full potential of the rich source domain knowledge to improve segmentation accuracy in unseen target domains. In this paper, we propose a source domain prior-assisted module for generalizable medical image segmentation based on SAM. Specifically, we store diverse features of the source domain data in a memory bank. When applying the model to the target domain, the features of the target domain first match with the features in the memory bank to obtain invaluable prior information. This strategy enables the model to utilize the prior information from the source domain to adapt to the target domain. We validate the proposed method on two widely used medical image segmentation tasks across multiple domains and achieve state-of-the-art performance.

Ethnobotanical study of medicinal plants of the Abohoy Gara Mountains, Northern Ethiopia

IntroductionFor many years, medicinal plants have been used in Ethiopia to manage a wide range of diseases. Documenting use of the plants for traditional medicine is necessary bases for obtaining new lead bioactive plant based natural products for the discovery of allopathic drugs, preserving the associated local knowledge and medicinal plants. This study aimed to document ethnobotanical knowledge associated with medicinal plants, parts used, route of preparations, and mode of administration. MethodsThe study was conducted from February 2017 to March 2018. The data were collected by interviewing 177 (139 male and 38 female) informants. Semistructured interviews, group discussions, guided field waks and quantitative approaches, such as informant consensus factors and ranking were used to collect and quantify the data. The data were analysed using Microsoft Excel, t test and analysis of variance (ANOVA). ResultsA total of 57 medicinal plant species belonging to 51 genera in 36 families were used by local people to prevent various ailments. The Asteraceae was the dominant family contributing the greatest number of medicinal plants. Leaves were the most widely used plant parts for treating human ailments (20, 56.7 %). Most of the collected traditional medicinal plants were used for treating human ailments directly (11, 28 %), while crushing was the most common method of preparation for treating livestock ailments (42, 70.2 %). Half of the herbal medicines were prepared for internal use, accounting for 31 (54.4 %). The findings showed that Cynoglossum coeruleum was the most preferred species for treating febrilosis, accounting for 33 cases. The highest informant consensus factor (0.83) was obtained for diseases related to febrilosis, and the lowest (0.56) was obtained for diseases associated with livestock (leech infestation and ectoparasites). ConclusionsThis study showed that medicinal plants in the study area, coupled with rich indigenous knowledge, were used by a large member of the population and are the most important means of treating some common ailments. However, medicinal plants in the study district are under pressure due to both natural and anthropogenic influences. Hence, the declining medicinal plants in the area require the application of complementary on-site and off-site conservation approaches.

Local Wisdom in Naming the Traditional Herbal Plants: A Semantic Analysis in Nias Language

This research investigates the local wisdom embedded in the naming of traditional herbal plants in the Nias language through a semantic analysis. The study addresses the gap in literature concerning the cultural and linguistic aspects of medicinal plant nomenclature in Nias, Indonesia. Utilizing a qualitative methodology, semi-structured interviews and documentation were conducted with five Nias individuals to gather insights on traditional herbal plants. Thirty traditional plant names were identified and analyzed for their semantic attributes. Findings reveal that the names often reflect physical characteristics, medicinal properties, or cultural significance, showcasing the rich botanical knowledge and cultural heritage of the Nias people. This study not only documents these unique plant names but also underscores the influence of local wisdom on the linguistic practices of the Nias community, highlighting the interplay between language, culture, and natural environment.

A self-attention TCN-based model for suicidal ideation detection from social media posts

Early suicidal ideation detection has long been regarded as an important task that can benefit both society and individuals. In this regard, it has been shown that, very frequently, the first symptoms of this problem can be identified by analyzing the contents shared on social media. Machine learning classification models have proven promising in capturing behavioral and textual features from posts shared on social media. This study proposes a novel machine-learning model to detect the risk of suicide from social media posts, employing both natural language processing and state-of-the-art deep learning techniques. We propose an ensemble LSTM-TCN model that benefits from a self-attention mechanism to detect suicidal ideation among users of two well-known social networks, Twitter (X) and Reddit. Furthermore, we present a comprehensive analysis of the data, examining the suicidal posts both statistically and semantically, which can provide rich knowledge about suicidal ideation. Our proposed model (AL-BTCN) outperforms the compared state-of-the-art models, resulting in over 94% accuracy, recall, and F1-score. Researchers, mental health specialists, and social media service providers can all benefit from the findings of this study.

Ethnomedicinal Uses of Plant Species in Panchase Protected Forests (A Case Study including the Kaski, Parbat & Syangja Districts of Nepal)

Ethnomedicinal plants, which have been utilized by indigenous communities in Nepal for centuries, hold promise for treating various ailments. Globally, medicinal plants are integral to healthcare systems, with Nepal renowned for its rich biodiversity and indigenous knowledge. This manuscript explores the ethno-medicinal practices within Nepal's Panchase protected forest, aiming to validate traditional remedies and contribute to healthcare improvement. Through a comprehensive study involving reconnaissance surveys, interviews, and data analysis, 103 ethno-medicinal plants from 58 families were documented, with Asteraceae being dominant. Various plant parts were utilized for treating 57 ailments, with leaves being most common. Gastrointestinal issues were prevalent, with plants such as Gheu kumari and Harro being frequently employed. Additionally, plants such as Dhursul and Ghodtapre were used for fever and migraine, while Khole Saag and Pani Amala addressed blood pressure and diabetes concerns. This study focuses on the importance of traditional medicine in rural healthcare and underscores the need for its preservation. Conservation efforts and intergenerational knowledge transfer initiatives are vital to safeguarding indigenous wisdom. Phytochemical and pharmacological studies are recommended to validate traditional knowledge scientifically. The commercial cultivation of medicinal plants could alleviate pressure on natural habitats and promote sustainable healthcare practices. This research contributes to the documentation of valuable ethno-medicinal knowledge, emphasizing its potential for socioeconomic development and healthcare enhancement in Nepal's rural communities.

Multi-Objective Evolutionary Neural Architecture Search with Weight-Sharing Supernet

Deep neural networks have played a crucial role in the field of deep learning, achieving significant success in practical applications. The architecture of neural networks is key to their performance. In the past few years, these architectures have been manually designed by experts with rich domain knowledge. Additionally, the optimal neural network architecture can vary depending on specific tasks and data distributions. Neural Architecture Search (NAS) is a class of techniques aimed at automatically searching for and designing neural network architectures according to the given tasks and data. Specifically, evolutionary-computation-based NAS methods are known for their strong global search capability and have aroused widespread interest in recent years. Although evolutionary-computation-based NAS has achieved success in a wide range of research and applications, it still faces bottlenecks in training and evaluating a large number of individuals during optimization. In this study, we first devise a multi-objective evolutionary NAS framework based on a weight-sharing supernet to improve the search efficiency of traditional evolutionary-computation-based NAS. This framework combines the population optimization characteristic of evolutionary algorithms with the weight-sharing ideas in one-shot models. We then design a bi-population MOEA/D algorithm based on the proposed framework to effectively solve the NAS problem. By constructing two sub-populations with different optimization objectives, the algorithm can effectively explore network architectures of various sizes in complex search spaces. An inter-population communication mechanism further enhances the algorithm’s exploratory capability, enabling it to find network architectures with uniform distribution and high diversity. Finally, we conduct performance comparison experiments on image classification datasets of different scales and complexities. Experimental results demonstrate the effectiveness of the proposed multi-objective evolutionary NAS framework and the practicality and transferability of the introduced bi-population MOEA/D-based NAS method compared to existing state-of-the-art NAS methods.

A meta analysis study on plants used by tribes of various districts of Odisha for treatment of liver disease

This review article include study on plants used by different tribes of different districts of Odisha for the treatment of Liver disease. A Field work was carried out the months of June – July 2016. The ethano-medicinal survey was conducted particularly with Odisha tribes in different districts having forest pockets with the help of local medicine men, locally known as Ayurveda. Data were collected through structure questionnaires and observing during the field visits. In the present study the local population used a total of 20 plant species belonging to 10 families to cure a variety of diseases. Out of the 20 plants,12 was herb, followed by trees and shrubs. The present study concluded that, the Odisha tribes of the study area possess rich knowledge on the medicinal plants and their utilization. Thus, the present study focuses on the documentation of the traditional knowledge of these valuable plants, which could enhance the potential of these medicinal plants to other communities as well and by understanding the importance, other communities can also helpful for conservation of these resources for further use.

身份認同文本對於有限或中斷正規教育成人學生的可供性

AbstractWorking with adult Students with Limited or Interrupted Formal Education (SLIFE) requires teachers to develop a knowledge and understanding of a complex range of matters related to pedagogy, language and literacy acquisition and the practical and socio-emotional challenges of resettlement. Limited research has focused on how teachers can support students’ literacy practices and the expression of students’ rich life experiences and knowledge systems through language learning. Using a design-based framework and undertaken across 10 weeks, this research investigated the pedagogical potential of identity texts when working with adult SLIFE in the Australian context. The collaborative research design involved teachers and teacher educators participating in ongoing professional learning sessions as a range of identity texts were integrated into lessons. Data collected included classroom recordings, recordings from professional learning sessions and formal interviews with teachers and materials produced by learners during the project. A thematic analysis of the data highlighted the transformative potential of the use of identity texts in creating more inclusive, student-centred learning environments and greater learner agency. Further work is needed to better understand the necessary pace, tempo and utility of a range of identity texts in adult SLIFE language learning contexts.

A single gene determines allorecognition in hydrozoan jellyfish Cladonema radiatum inbred lines.

Allorecognition-the ability of an organism to discriminate between self and nonself-is crucial to colonial marine animals to avoid invasion by other individuals in the same habitat. The cnidarian hydroid Hydractinia has long been a major research model in studying invertebrate allorecognition, establishing a rich knowledge foundation. In this study, we introduce a new cnidarian model Cladonema radiatum (C. radiatum). C. radiatum is a hydroid jellyfish which also forms polyp colonies interconnected with stolons. Allorecognition responses-fusion or regression of stolons-are observed when stolons encounter each other. By transmission electron microscopy, we observe rapid tissue remodeling contributing to gastrovascular system connection in fusion. Meanwhile, rejection responses are regulated by reconstruction of the chitinous exoskeleton perisarc, and induction of necrotic and autophagic cellular responses at cells in contact with the opponent. Genetic analysis identifies allorecognition genes: six Alr genes located on the putative allorecognition complex and four immunoglobulin superfamily genes on a separate genome region. C. radiatum allorecognition genes show notable conservation with the Hydractinia Alr family. Remarkedly, stolon encounter assays of inbred lines reveal that genotypes of Alr1 solely determine allorecognition outcomes in C. radiatum.

Application of machine reading comprehension techniques for named entity recognition in materials science

Materials science is an interdisciplinary field that studies the properties, structures, and behaviors of different materials. A large amount of scientific literature contains rich knowledge in the field of materials science, but manually analyzing these papers to find material-related data is a daunting task. In information processing, named entity recognition (NER) plays a crucial role as it can automatically extract entities in the field of materials science, which have significant value in tasks such as building knowledge graphs. The typically used sequence labeling methods for traditional named entity recognition in material science (MatNER) tasks often fail to fully utilize the semantic information in the dataset and cannot effectively extract nested entities. Herein, we proposed to convert the sequence labeling task into a machine reading comprehension (MRC) task. MRC method effectively can solve the challenge of extracting multiple overlapping entities by transforming it into the form of answering multiple independent questions. Moreover, the MRC framework allows for a more comprehensive understanding of the contextual information and semantic relationships within materials science literature, by integrating prior knowledge from queries. State-of-the-art (SOTA) performance was achieved on the Matscholar, BC4CHEMD, NLMChem, SOFC, and SOFC-Slot datasets, with F1-scores of 89.64%, 94.30%, 85.89%, 85.95%, and 71.73%, respectively in MRC approach. By effectively utilizing semantic information and extracting nested entities, this approach holds great significance for knowledge extraction and data analysis in the field of materials science, and thus accelerating the development of material science.Scientific contributionWe have developed an innovative NER method that enhances the efficiency and accuracy of automatic entity extraction in the field of materials science by transforming the sequence labeling task into a MRC task, this approach provides robust support for constructing knowledge graphs and other data analysis tasks.

Joint Dual Feature Distillation and Gradient Progressive Pruning for BERT compression

The increasing size of pre-trained language models has led to a growing interest in model compression. Pruning and distillation are the primary methods employed to compress these models. Existing pruning and distillation methods are effective in maintaining model accuracy and reducing its size. However, they come with limitations. For instance, pruning is often suboptimal and biased by transforming it into a continuous optimization problem. Distillation relies primarily on one-to-one layer mappings for knowledge transfer, which leads to underutilization of the rich knowledge in teacher. Therefore, we propose a method of joint pruning and distillation for automatic pruning of pre-trained language models. Specifically, we first propose Gradient Progressive Pruning (GPP), which achieves a smooth transition of indicator vector values from real to binary by progressively converging the values of unimportant units’ indicator vectors to zero before the end of the search phase. This effectively overcomes the limitations of traditional pruning methods while supporting compression with higher sparsity. In addition, we propose the Dual Feature Distillation (DFD). DFD adaptively globally fuses teacher features and locally fuses student features, and then uses the dual features of global teacher features and local student features for knowledge distillation. This realizes a “preview-review” mechanism that can better extract useful information from multi-level teacher information and transfer it to student. Comparative experiments on the GLUE benchmark dataset and ablation experiments indicate that our method outperforms other state-of-the-art methods.

서비스형 버추얼 휴먼의 감성적 디자인 분석 및 UX평가

As the digitization of museums accelerates, service-oriented virtual humans are leading the museum industry's entry into the metaverse. However, compared to other countries, the application of service-oriented virtual humans in Chinese museums is still in its early stages. This study explores ways to enhance the user experience of service-oriented virtual humans in museums through emotional design and UX evaluation. Firstly, the study identifies emotional design elements and UX evaluation factors through prior research. Secondly, it discusses seven elements—visual, auditory, storytelling, interaction, pleasure, satisfaction, and convenience—across the three levels of emotional design(instinctive, behavioral, and reflective) using representative cases from Chinese museums. Thirdly, based on prior research, the study establishes UX evaluation criteria and selects expert interview subjects, identifying key factors affecting user experience: meaningful, pleasurable, convenient, usable, reliable, and interaction. Finally, the study synthesizes insights from case analysis and interview results. From an emotional design perspective, virtual humans are modeled after real people and incorporate traditional Chinese cultural elements in their appearance, using natural dialogue and emotional expression. They explain the background and history of artifacts and enhance participation through diverse interaction methods. Rich knowledge and engaging interactions attract visitor attention and enhance enjoyment. In terms of UX evaluation, diverse interactions increase immersion, while natural language and visual/auditory elements stimulate curiosity and engagement. Concise information delivery and voice interaction improve convenience. Functions such as inquiry response and route planning meet individual needs. Natural voice interaction and data security enhance trust, while improved voice recognition accuracy and facial expression consistency enhance user immersion and interaction experience.

MOTIF FEATURES OF THE WORKS «KOKSEREK»BY MUKHTAR AUEZOVAND «WHITE FANG» BY JACK LONDON

«Kokserek»and «White Fang»are stories written by the famous American novelist Jack London and the famous Kazakh writer Mukhtar Auezov. It is obvious that the clear depiction of the image of a dog and a wolf with a convincing, artistic accuracy in a truly artistic language requiresa great creative search from the writer. First, it is due to the writer’s rich knowledge gained from the best examples of the world literature, and secondly, the natural talent and his own life experience. The present paper identifies the motiffeatures in the works "Kokserek" and "White Fang". The study relevance lies in the identification of two facts. The first fact is that the image of “wolf” cannot be found in the Kazakh prose before the work “Kokserek” by M. Auezov although this phenomenon had already found its place in world literature. Secondly, it is known that Mukhtar Auezov translated the work "Volk" ("Wolf") by Jack London in 1936. This simply explains the creation of a complex image of "Kokserek" by the great writer and the presence of harmony and originality inherent in both works. The study purpose is to conduct a comparative analysis on the plot features of the works. The works "Kokserek" by Mukhtar Auezov and "White Fang" by Jack London served as the material for the study. The study results shows that the metaphorical image of “wolf” created by Jack London later developed, became more complex, recognizable and supplemented in a new way in “Kokserek” of M. Auezov. Theoretical and practical significance of the work is that the study result can be used as a methodological tool for comparing the works of world literature. The scientific novelty of the study presents a number of particular studies and conclusions based on the analysis of specific materials.