Research on machine assisted text analysis follows the rapid development of digital media, and sentiment analysis is among the prevalent applications. Traditional sentiment analysis methods require ...complex feature engineering, and embedding representations have dominated leaderboards for a long time. However, the context-independent nature limits their representative power in rich context, hurting performance in Natural Language Processing (NLP) tasks. Bidirectional Encoder Representations from Transformers (BERT), among other pre-trained language models, beats existing best results in eleven NLP tasks (including sentence-level sentiment classification) by a large margin, which makes it the new baseline of text representation. As a more challenging task, fewer applications of BERT have been observed for sentiment classification at the aspect level. We implement three target-dependent variations of the BERT base model, with positioned output at the target terms and an optional sentence with the target built in. Experiments on three data collections show that our TD-BERT model achieves new state-of-the-art performance, in comparison to traditional feature engineering methods, embedding-based models and earlier applications of BERT. With the successful application of BERT in many NLP tasks, our experiments try to verify if its context-aware representation can achieve similar performance improvement in aspect-based sentiment analysis. Surprisingly, coupling it with complex neural networks that used to work well with embedding representations does not show much value, sometimes with performance below the vanilla BERT-FC implementation. On the other hand, incorporation of target information shows stable accuracy improvement, and the most effective way of utilizing that information is displayed through the experiment.
Abstract
Many high quality studies have emerged from public databases, such as Surveillance, Epidemiology, and End Results (SEER), National Health and Nutrition Examination Survey (NHANES), The ...Cancer Genome Atlas (TCGA), and Medical Information Mart for Intensive Care (MIMIC); however, these data are often characterized by a high degree of dimensional heterogeneity, timeliness, scarcity, irregularity, and other characteristics, resulting in the value of these data not being fully utilized. Data-mining technology has been a frontier field in medical research, as it demonstrates excellent performance in evaluating patient risks and assisting clinical decision-making in building disease-prediction models. Therefore, data mining has unique advantages in clinical big-data research, especially in large-scale medical public databases. This article introduced the main medical public database and described the steps, tasks, and models of data mining in simple language. Additionally, we described data-mining methods along with their practical applications. The goal of this work was to aid clinical researchers in gaining a clear and intuitive understanding of the application of data-mining technology on clinical big-data in order to promote the production of research results that are beneficial to doctors and patients.
Abstract Cancer is a complex disease composing systemic alterations in multiple scales. In this study, we develop the Tumor Multi-Omics pre-trained Network (TMO-Net) that integrates multi-omics ...pan-cancer datasets for model pre-training, facilitating cross-omics interactions and enabling joint representation learning and incomplete omics inference. This model enhances multi-omics sample representation and empowers various downstream oncology tasks with incomplete multi-omics datasets. By employing interpretable learning, we characterize the contributions of distinct omics features to clinical outcomes. The TMO-Net model serves as a versatile framework for cross-modal multi-omics learning in oncology, paving the way for tumor omics-specific foundation models.
The technology of automatic text generation by machine has always been an important task in natural language processing, but the low-quality text generated by the machine seriously affects the user ...experience due to poor readability and fuzzy effective information. The machine-generated text detection method based on traditional machine learning relies on a large number of artificial features with detection rules. The general method of text classification based on deep learning tends to the orientation of text topics, but logical information between texts sequences is not well utilized. For this problem, we propose an end-to-end model which uses the text sequences self-information to compensate for the information loss in the modeling process, to learn the logical information between the text sequences for machine-generated text detection. This is a text classification task. We experiment on a Chinese question and answer the dataset collected from a biomedical social media, which includes human-written text and machine-generated text. The result shows that our method is effective and exceeds most baseline models.
Osteoporosis is becoming a common skeletal disorder characterised by reduced bone strength that leads to an increased risk for fractures. Despite its increasing prevalence among the Chinese ...population, osteoporosis remains both underdiagnosed and undertreated. Many observational studies have reported prevalence of osteoporosis in China; however, the estimated prevalence varies within a certain range because of the diverse ways of life and diverse genetic heritage of the populations in different Chinese regions, and few studies have examined the sex-specific prevalence of osteoporosis in China across time. We did a systematic review and meta-analysis to compare the prevalence and temporal trends of osteoporosis in Chinese women and men.
In accordance with MOOSE and PRISMA, we did literature searches in PubMed, MEDLINE, Embase, the China National Knowledge Infrastructure, and the Chinese WanFang databases for epidemiological studies that reported osteoporosis prevalence in Chinese individuals from their inception to Jan 31, 2020, without language restrictions. Studies were considered eligible if they reported sufficiently detailed results about prevalence of osteoporosis among Chinese individuals aged at least 18 years, and if they used either WHO or Chinese criteria for osteoporosis diagnosis. Studies were excluded if they were reviews, case reports, editorials, guidelines, and other article forms without available full-texts, if they were duplicate papers or evaluating the same sample, and if the studies were done in a population with other specific diseases. We assessed prevalence of osteoporosis using a random-effects model by applying binomial distribution and Freeman-Tukey double arcsine transformation to stabilise the variances, and the pooled Freeman-Tukey prevalence estimates were then back transformed to their original scale for ease of interpretation. We also did a subgroup analysis by sex, age, region, and screening period.
We identified 129 eligible studies that included 982 563 individuals. The overall prevalence of osteoporosis in China was 20·80% (95% CI 17·95–23·79). Its prevalence was higher in women (23·57%, 18·50–29·04) than in men (12·22%, 7·23–18·29), and in rural (23·06%, 16·80–29·99) than urban regions (20·95%, 17·09–25·09). From 2001–05 to 2016–19, prevalence of osteoporosis progressed steadily from 19·35% (14·86–24·28) to 21·30% (17·17–25·73) in 2006–10 and 22·61 (19·27–26·12) to 23·58% (17·08–30·78) in 2011–15. In terms of temporal stratification, the prevalence increased with age in both sexes, with a greater rise in women from age 50–59 (10·22%) years to 80 years and older (62·24%) than in men of the same age groups (from 5·62% to 30·55%). In terms of regional stratification, the pooled prevalence estimates were higher among Chinese individuals in the south, north, and southwest (prevalence of more than 25%) by a factor of 1·5, compared with their counterparts in the northeast of China. Regional prevalence of osteoporosis ranged from 25·78% (19·79–32·26) in the south, 22·55% (14·32–32·02) in the east, 19·60% (13·25–26·85) in the northwest, and 17·33% (11·23–24·44) in the northeast.
This up-to-date meta-analysis provides a comprehensive overview of the burden of osteoporosis among Chinese women and men. Our findings suggest a considerable prevalence of osteoporosis, especially in older Chinese women. Our findings also emphasise the urgent need for control measures and preventive management of osteoporosis because of a rapidly ageing population in China.
National Natural Science Foundation of China (81803329) and China Postdoctoral Science Foundation (2018M631780).
Multilabel classification is one of the most challenging tasks in natural language processing, posing greater technical difficulties than single-label classification. At the same time, multilabel ...classification has more natural applications. For individual labels, the whole piece of text has different focuses or component distributions, which require full use of local information of the sentence. As a widely adopted mechanism in natural language processing, attention becomes a natural choice for the issue. This paper proposes a multilayer self-attention model to deal with aspect category and word attention at different granularities. Combined with the BERT pretraining model, it achieves competitive performance in aspect category detection and electronic medical records’ classification.
Electrochemical water splitting to produce green hydrogen offers a promising technology for renewable energy conversion and storage, as well as realizing carbon neutrality. The efficiency, stability, ...and cost of electrocatalysts toward hydrogen evolution reaction (HER) and electrocatalytic overall water splitting (EOWS) at large current densities are essential for practical application. In this review, the key factors that determine the catalytic performance of electrocatalysts at large current densities are summarized from the angel of thermodynamic and kinetic correlation. The corresponding design strategies are presented. The electronic structure and density of active sites that affect the adsorption/desorption of intermediates are considered as the thermodynamic aspects, while charge transfer and mass transport capabilities closely associated with electrode resistance and intermediate diffusion are assigned as kinetic effects. Recent development of bifunctional and integrated electrocatalysts toward EOWS is also discussed in detail. Finally, the perspective and direction on the electrocatalytic water splitting under large current density are proposed. This comprehensive overview will offer profound insights and guidance for the continued advancement of this field.
This review discusses the key factors that determine the hydrogen evolution performance of electrocatalysts at large current densities from the angle of thermodynamic and kinetic correlation, as well as presents perspective and future direction, which will provide profound insights and guidance for the continued advancement of this field.
The objectives of the survey were to identify the level of influenza vaccination coverage in China in three influenza seasons 2009/10 to 2011/12, and to find out potential predictors for seasonal ...influenza vaccination.
In September and October 2011, representative urban household telephone surveys were conducted in five provinces in China with a response rate of 6%. Four target groups were defined for analysis: 1) children ≤ 5 years old; 2) elderly persons aged ≥ 60 years old; 3) health care workers (persons working in the medical field) and 4) chronically ill persons.
The overall mean vaccination rate was 9.0%. Among the four target groups, the rate of vaccination of children aged ≤ 5 years old (mean = 26%) was highest and the rate of elderly people aged ≥ 60 years old (mean = 7.4%) was the lowest, while the rates of persons who suffer from a chronic illness (mean = 9.4%) and health care workers (9.5%) were similar. A subsidy for influenza vaccination, age group, health care workers, suffering from a chronic illness and living in Eastern China were independent significant predictors for influenza vaccination.
The seasonal influenza vaccination coverage rates among urban populations in selected cities and provinces in China were far below previously reported rates in developed countries. Influenza vaccination coverage rates differed widely between different target groups and provinces in China. Subsidy policy might have a positive effect on influenza vaccination rate, but further cost-effectiveness studies, as well as the vaccination rate associated factors studies are still needed to inform strategies to increase coverage.
Nanomedicines confront various complicated physiological barriers limiting the accumulation and deep penetration in the tumor microenvironment, which seriously restricts the efficacy of antitumor ...therapy. Self‐propelled nanocarriers assembled with kinetic engines can translate external energy into orientated motion for tumor penetration. However, achieving a stable ultrafast permeability at the tumor site remains challenging. Here, sub‐200 nm photoactivated completely organic nanorockets (NRs), with asymmetric geometry conveniently assembled from photothermal semiconducting polymer payload and thermo‐driven macromolecular propulsion through a straightforward nanoprecipitation process, are presented. The artificial NRs can be remotely manipulated by 808 nm near‐infrared light to trigger the photothermal conversion and Curtius rearrangement reaction within the particles for robustly pushing nitrogen out into the solution. Such a two‐stage light‐to‐heat‐to‐chemical energy transition effectively powers the NRs for an ultrafast (≈300 µm s−1) and chemical medium‐independent self‐propulsion in the liquid media. That endows the NRs with high permeability against physiological barriers in the tumor microenvironment to directionally deliver therapeutic agents to target lesions for elevating tumor accumulation, deep penetration, and cellular uptake, resulting in a significant enhancement of antitumor efficacy. This work will inspire the design of advanced kinetic systems for powering intelligent nanomachines in biomedical applications.
Engineered organic nanorockets are powered by photoactivated organic kinetic systems through a two‐stage light‐to‐heat‐to‐chemical energy transition for a stable ultrafast (≈300 µm s−1) self‐propulsion in the liquid media. The programmable navigation allows a high permeability against physiological barriers for elevating accumulation and deep penetration at the tumor site, thereby significantly enhancing the antitumor efficacy of the nanomedicines.
Limited permeability in solid tumors significantly restricts the anticancer efficacy of nanomedicines. Light‐driven nanomotors powered by photothermal converting engines are appealing carriers for ...directional drug delivery and simultaneous phototherapy. Nowadays, it is still a great challenge to construct metal‐free photothermal nanomotors for a programmable anticancer treatment. Herein, one kind of photoactivated organic nanomachines is reported with asymmetric geometry assembled by light‐to‐heat converting semiconducting polymer engine and macromolecular anticancer payload through a straightforward nanoprecipitation process. The NIR‐fueled polymer engine can be remotely controlled to power the nanomachines for light‐driven thermophoresis in the liquid media and simultaneously thermal ablating the cancer cells. The great manipulability of the nanomachines allows for programming of their self‐propulsion in the tumor microenvironment for effectively improving cellular uptake and tumor penetration of the anticancer payload. Taking the benefit from this behavior, a programmed treatment process is established at a low drug dose and a low photothermal temperature for significantly enhancing the antitumor efficacy.
Photoactivated organic nanomachines are powered by an NIR‐fueled organic semiconducting polymer engine for light‐driven traversing of physiological barriers in the tumor microenvironment. The great manipulability of the organic nanomachines allows precise programming of the antitumor treatment for significantly enhanced efficacy at a low drug dose and a low photothermal temperature.