检索结果-内蒙古大学图书馆

Practical Applications of ChatGPT and BERT: Long Text Summarization and Model Feature Ablation Study for Stock Comment Topic Identification 24

Practical Applications of ChatGPT and BERT: Long Text Summar...

引用

2024 Guangdong-Hong Kong-Macao Greater Bay Area International Conference on Digital Economy and Artificial Intelligence, DEAI 2024

作者： Xu, Mingyan chen, Hongyu Deng, fuwen Liu, Yuan Zhao, Feng Zhang, Yong School of Statistics Shandong Technology and Business University Yantai264000 China School of Hospital Shandong Technology and Business University Yantai264000 China School of Computer Science and Technology Shandong Technology and Business University Yantai264000 China Yantai Key Laboratory of Big Data Modeling and Intelligent Computing School of Computer Science and Technology Shandong Technology and Business University Yantai China School of Information and Electrotic Shandong Technology and Business University Yantai264000 China

ISBN: (纸本)9798400717147

Topic identification of expert stock comments is an important task in the field of finance, which helps investors understand the market dynamics and expert opinions so that they can make more informed investment decisions. The Bidirectional Encoder Representations from Transformers (BERT) model is one of the widely used methods for text topic identification. To improve the identification accuracy and overcome the limitations of the BERT model, two strategies are proposed in this study. First, given the BERT model's input length limitation when dealing with long texts, we used the summaries created by Chat Generative Pre-training Transformer (ChatGPT) as inputs to the model. The generative summarization capability of ChatGPT not only reduces the length of the text effectively but also preserves the key information of the original text, which enables the model to better understand and deal with long texts. Second, we perform multi-layer feature ablation on the BERT model to find the most representative features for text topic identification. The multi-layer features of the BERT model can be more effectively utilized to improve the accuracy of text topic identification. Experiments show that our proposed strategies exhibit excellent performance and provide effective paths to improve the accuracy of stock comment topic identification. © 2024 ACM.

关键词： Ablation

来源：评论

学校读者我要写书评

暂无评论

The genomic landscape of short tandem repeats in cattle

引用

ANIMAL GENETICS 2025年第1期56卷 e13498页

作者： Wang, Pengfei Sheng, Xin Xia, Xiaoting Wang, fuwen Li, Ruizhe Ahmed, Zulfiqar chen, Ningbo Lei, Chuzhao Ma, Zhijie Qinghai Univ State Key Lab Plateau Ecol & Agr Xining 810016 Qinghai Peoples R China Northwest A&F Univ Coll Anim Sci & Technol Key Lab Anim Genet Breeding & Reprod Shaanxi Prov Yangling 712100 Shaanxi Peoples R China Qinghai Univ Acad Anim Sci & Vet Med Xining Peoples R China Univ Poonch Rawalakot Fac Vet & Anim Sci Dept Livestock & Poultry Prod Azad Jammu & Kashmir Pakistan

Short tandem repeats (STRs) are abundant and have high mutation rates across cattle genomes;however, comprehensive exploration of cattle STRs is needed. Here, we constructed a comprehensive map of 467 553 polymorphic STRs (pSTRs) constructed from 423 cattle genomes representing 59 breeds worldwide. We observed that pSTRs in coding sequences and 5 ' UTRs (Untranslated Regions) were under strong selective constraints and exhibited a relatively low level of diversity. Furthermore, we found that these pSTRs underwent more contraction than expansion. Population analysis showed a strong positive correlation (R = 1) between pSTR diversity and single nucleotide polymorphic heterozygosity. We also investigated STR differences between taurine and indicine cattle and detected 2301 highly divergent STRs, which might relate to immune, endocrine and neurodevelopmental pathways. In summary, our large-scale study characterizes the spectrum of STRs in cattle, expands the scale of known cattle STR variation and provides novel insights into differences among various cattle subspecies.

关键词： cattle diversity population genetics selection pressure STR

来源：评论

学校读者我要写书评

暂无评论

The influence of aging time and moisture content on the tensile strength of transformer bushing insulating paper

引用

AIP ADVANCES 2022年第11期12卷 1-6页

作者： Dai, Quanmin Mu, Guowei Tan, fuwen chen, Huidong You, Xianggang Liu, Yanxia cheng, Guang Beijing Union Univ Coll Urban Rail Transit & Logist Beijing 100101 Peoples R China

The parameter of tensile strength for transformer bushing insulating paper occupies an important position in the structural design and life evaluation of transformer bushing. In order to precisely achieve the effect of aging time and moisture content on the tensile strength of transformer bushing paper, transformer insulating paper samples with a thickness of 0.13 mm were prepared for tensile strength and dielectric strength testing under different moisture contents and aging states. The results show that the tensile strength (MPa) of insulating paper decreases linearly in the range of 0-648 h under accelerated thermal aging at 130 & DEG;C. However, there is a more serious exponential decline in the tensile strength (MPa) of insulation paper with its moisture content is boosted. The moisture content in insulation paper is the most important factor for insulation paper deterioration in oil-impregnated paper insulation equipment, which could accelerate the end of insulation service life. The value of capacitance increment & UDelta;C of the insulation paper samples displayed closely linear growth with a moisture content from 0.5% to 5.9%. Strictly controlling the moisture content by monitoring the delta C value of transformer insulation paper is of significance for prolonging its service life. (C) 2022 Author(s).All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY)license (http://***/licenses/by/4.0/).

关键词： Tensile strength

来源：评论

学校读者我要写书评

暂无评论

Performance evaluation of a deep learning-based cascaded HRNet model for automatic measurement of X-ray imaging parameters of lumbar sagittal curvature

引用

EUROPEAN SPINE JOURNAL 2024年第11期33卷 4104-4118页

作者： Wu, Yuhua chen, Xiaofei Dong, fuwen He, Linyang cheng, Guohua Zheng, Yuwen Ma, Chunyu Yao, Hongyan Zhou, Sheng Gansu Univ Chinese Med Clin Med Coll 1 Lanzhou 730000 Gansu Peoples R China Gansu Univ Tradit Chinese Med Gansu Prov Hosp Tradit Chinese Med Affiliated Hosp 1 Dept Radiol Lanzhou 730050 Gansu Peoples R China Hangzhou Jianpei Technol Co Ltd Hangzhou 311200 Zhejiang Peoples R China Gansu Prov Hosp Dept Radiol 204 Donggang West Rd Lanzhou 730000 Gansu Peoples R China

PurposeTo develop a deep learning-based cascaded HRNet model, in order to automatically measure X-ray imaging parameters of lumbar sagittal curvature and to evaluate its prediction *** total of 3730 lumbar lateral digital radiography (DR) images were collected from picture archiving and communication system (PACS). Among them, 3150 images were randomly selected as the training dataset and validation dataset, and 580 images as the test dataset. The landmarks of the lumbar curve index (LCI), lumbar lordosis angle (LLA), sacral slope (SS), lumbar lordosis index (LLI), and the posterior edge tangent angle of the vertebral body (PTA) were identified and marked. The measured results of landmarks on the test dataset were compared with the mean values of manual measurement as the reference standard. Percentage of correct key-points (PCK), intra-class correlation coefficient (ICC), Pearson correlation coefficient (r), mean absolute error (MAE), mean square error (MSE), root-mean-square error (RMSE), and Bland-Altman plot were used to evaluate the performance of the cascade HRNet *** PCK of the cascaded HRNet model was 97.9-100% in the 3 mm distance threshold. The mean differences between the reference standard and the predicted values for LCI, LLA, SS, LLI, and PTA were 0.43 mm, 0.99 degrees, 1.11 degrees, 0.01 mm, and 0.23 degrees, respectively. There were strong correlation and consistency of the five parameters between the cascaded HRNet model and manual measurements (ICC = 0.989-0.999, R = 0.991-0.999, MAE = 0.63-1.65, MSE = 0.61-4.06, RMSE = 0.78-2.01).ConclusionThe cascaded HRNet model based on deep learning algorithm could accurately identify the sagittal curvature-related landmarks on lateral lumbar DR images and automatically measure the relevant parameters, which is of great significance in clinical application.

关键词： Deep learning Automatic measurement Lumbar sagittal curvature Lumbar lordosis angle (LLA) Sacral slope (SS) Radiography

来源：评论

学校读者我要写书评

暂无评论

Conservation genomics of the critically endangered Chinese pangolin

引用

Science China(Life Sciences) 2024年第10期67卷 2051-2061页

作者： Shichao Wei Huizhong Fan Wenliang Zhou Guangping Huang Yan Hua Shibao Wu Xiao Wei Yiting chen Xinyue Tan fuwen Wei Jiangxi Province Key Laboratory of Conservation Biology College of ForestryJiangxi Agricultural UniversityNanchang 330045China Center for Evolution and Conservation Biology Southern Marine Science and Engineering Guangdong Laboratory(Guangzhou)Guangzhou 511458China Institute of Zoology Chinese Academy of SciencesBeijing 100101China Guangdong Provincial Key Laboratory of Silviculture Protection and UtilizationGuangdong Academy of ForestryGuangzhou 510520China School of Life Science South China Normal UniversityGuangzhou 510631China Terrestrial Wildlife Rescue and Epidemic Diseases Surveillance Center of Guangxi Nanning 530003China

The Chinese pangolin(Manis pentadactyla,MP)has been extensively exploited and is now on the brink of extinction,but its population structure,evolutionary history,and adaptive potential are ***,we analyzed 94 genomes from three subspecies of the Chinese pangolin and identified three distinct genetic clusters(MPA,MPB,and MPC),with MPB further divided into MPB1 and MPB2 *** divergence of these populations was driven by past climate *** MPB2 and MPC,recent human activities have caused dramatic population decline and small population size as well as increased inbreeding,but not decrease in genomic variation and increase in genetic load probably due to strong gene flow;therefore,it is crucial to strengthen in situ habitat management for these two *** contrast,although human activities have a milder impact on MPA,it is at high risk of extinction due to long-term contraction and isolation,and genetic rescue is urgently ***1 exhibited a relatively healthy population status and can potentially serve as a source ***,our findings provide novel insights into the conservation of the Chinese pangolin and biogeography of the mammals of eastern Asia.

关键词： Manis pentadactyla conservation genomics demographic history human activities conservation strategies

来源：评论

学校读者我要写书评

暂无评论

Network pharmacology and experimental validation to explore the role and potential mechanism of Liuwei Dihuang Decoction in prostate cancer

引用

BMC COMPLEMENTARY MEDICINE AND THERAPIES 2024年第1期24卷 284页

作者： Zhan, Xiangyang Li, Haoze Jin, Jingyun Ju, Xiran Gao, Jiawei chen, Xinglin Yuan, fuwen Gu, Jianyi Xu, Dongliang Ju, Guanqun Shanghai Univ Tradit Chinese Med Urol Ctr Shuguang Hosp Shanghai 201203 Peoples R China Shanghai Univ Tradit Chinese Med Surg Inst Integrat Med Shuguang Hosp Shanghai 201203 Peoples R China Shanghai Univ Tradit Chinese Med Surg Inst Shuguang Hosp Shanghai 201203 Peoples R China Shanghai Univ Tradit Chinese Med Shanghai 201203 Peoples R China

ObjectiveTo evaluate the anti-tumor effector of Liuwei Dihuang Decoction (LWDHD) in prostate cancer (PCa) and explore the potential mechanism using experimental validation, network pharmacology, bioinformatics analysis, and molecular *** test, Clone formation assay and wound-healing assays were used to determine the effect of LWDHD on prostate cancer growth and metastasis. The active ingredients and targets of LWDHD were obtained from the TCMSP database, and the relevant targets were selected by GeneCards, OMIM and DisGeNET databases for PCa. The cross-targets of drugs and disease were imported into the STRING database to construct protein interactions. The network was also visualized using Cytoscape software and core targets are screened using the Network Analyzer plug-in. The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment were analyzed using R software. TCGA database was used to analyze the correlation of bioinformatics genes. AutoDock vina was used to predict the molecular docking and binding ability of active ingredients to key targets. Through WB and q-PCR experiments, the above gene targets were detected to verify the effect of LWDHD on *** and scratch tests confirmed that LWDHD could inhibit the proliferation, invasion and migration of prostate cancer cells. Clone formation experiments showed that LWDHD inhibited the long-term proliferative capacity of PC3 cells. LWDHD and PCa had a total of 99 common targets, establishing a "drug-ingredient-common target" network. Through GO and KEGG enrichment analysis, PI3K/AKT, MAPK, TP53 pathway, MYC, TNF pathway and other signaling pathways were found. Bioinformatics analysis showed that MYC gene was highly expressed and CCND1 and MAPK1 were low expressed in prostate cancer tissues. In addition, TP53, AKT1, MYC, TNF and CCND1 were positively correlated with MAPK1, among which AKT1 and CCND1 were most closely correlated with MAPK1. Molecular docking resu

关键词： Liuwei Dihuang Decoction Prostate cancer Network pharmacology Bioinformation analysis Experimental validation

来源：评论

学校读者我要写书评

暂无评论

Characteristics, Source Apportionment, and Health Risk of Heavy Metals in the Soils of Peri-urban Shanghai Chongming Island

引用

ACS OMEGA 2024年第42期9卷 42734-42745页

作者： Shen, cheng Wang, Min Su, Jinghua Sun, Huilun Hu, Wenan Lin, Kuangfei Wu, Jian Liu, fuwen chen, Xiurong Sha, chenyan East China Univ Sci & Technol Sch Resources & Environm Engn State Environm Protect Key Lab Environm Risk Asses Shanghai 200237 Peoples R China Shanghai Acad Environm Sci State Environm Protect Engn Ctr Urban Soil Contami Shanghai 200233 Peoples R China East China Normal Univ Sch Ecol & Environm Sci Shanghai 200241 Peoples R China Shanghai Technol Ctr Reduct Pollut & Carbon Emiss Shanghai 200235 Peoples R China Shanghai Inst Technol Sch Chem & Environm Engn Shanghai 201418 Peoples R China

Heavy metals resulting from human activities pose significant threats to human health and the soil ecosystem. In the current study, 917 soil samples from Chongming Island in Shanghai, China, were examined for eight heavy metals. The sources of contamination were identified by using a Positive Matrix Factorization (PMF) model. Meanwhile, spatial interpolation and Moran's I index were applied to validate the model in terms of spatial linkages. The results revealed that the average concentrations of As, Cd, Hg, Pb, Cr, Cu, Zn, and Ni in the soil were 8.87, 0.19, 0.06, 28.75, 76.01, 37.74, 88.93, and 30.33 mg kg(-1), respectively. The PMF analysis proved that heavy metals in the soil of the study area are mainly influenced by traffic sources (Cr and Pb), industrial sources (Zn, Cd, and Cu), station sources (Hg), and natural sources (As and Ni), with contribution rates of 22.23, 26.25, 36.38, and 15.14%, respectively. The combination of Moran's index and the spatial analysis method not only verified the analytical results of the receptor model on the one hand but also served as a supplementary explanation for the sources of heavy metals in the soil. The health risk assessment indicated that noncarcinogenic values were below the threshold values. The total carcinogenic risk (R-T) of different heavy metals has a descending order of Cr > As > Ni > Cd. The R-T values of multiple heavy metals for children and adults were 5.28 x 10(-04) and 4.10 x 10(-05), respectively, which were close to the risk threshold. Therefore, attention should be paid to the health risks, especially for children's skin contact, which is the main exposure pathway.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A high-temperature-resistant die attach material based on Cu@In@Ag particles for high-power devices

引用

JOURNAL OF MATERIALS SCIENCE-MATERIALS IN ELECTRONICS 2022年第8期33卷 5599-5612页

作者： Mao, Xingchao Liu, Jiahao Yu, fuwen Fu, Xing Hang, Chunjin chen, Hongtao Li, Mingyu Harbin Inst Technol Shenzhen Sauvage Lab Smart Mat Shenzhen 518055 Peoples R China Harbin Inst Technol State Key Lab Adv Welding & Joining Harbin 150001 Peoples R China China Elect Product Reliabil & Environm Testing R Sci & Technol Reliabil Phys & Applicat Elect Comp Guangzhou 510640 Peoples R China

This paper presents a high-temperature-resistant preform based on Cu@In@Ag particles that can be used as a die attach material for high-power devices, using an electroless method to achieve the preparation of core-shell structure particles. Indium has a lower melting point of 156.6 degrees C and can be completely consumed to form intermetallic compounds of Cu2In, Ag9In4, and Ag3In, which will raise the melting point of the solder joint to over 400 degrees C. The silver coating is intended to enhance the oxidation resistance of the Cu@In particles and to improve the service reliability of the solder joint. It was demonstrated that high-temperature-resistant solder joints could be obtained by reflowing at 200 degrees C for 15 min under a pressure of 10 MPa. The shear strengths of solder joints were 29.70 MPa and 21.85 MPa at 25 degrees C and 300 degrees C, respectively. The thermal and electrical properties of the solder joints after reflow composed of Cu@In@Ag core-shell particles were tested and the thermal conductivity was 79.02 W/m K and 58.64 W/m K at 30 degrees C and 250 degrees C, respectively. Great electrical and thermal conductivity and high shear strength at high temperatures result in high-temperature reliability. As a high-temperature-resistant die attachment material, it has the potential to be used in die attachment in high-power equipment modules.

关键词： Silver alloys

来源：评论

学校读者我要写书评

暂无评论

Assessing the response of marine fish communities to climate change and fishing

引用

CONSERVATION BIOLOGY 2024年第6期38卷 e14291页

作者： Huang, Mingpan chen, Yiting Zhou, Wenliang Wei, fuwen Ctr Evolut & Conservat Biol Southern Marine Sci & Engn Guangdong Lab Guangzhou 1119 Haibin Rd Guangzhou 511458 Peoples R China Inst Zool Chinese Acad Sci CAS Key Lab Anim Ecol & Conservat Biol Beijing Peoples R China Hong Kong Univ Sci & Technol Dept Ocean Sci Hong Kong Peoples R China Jiangxi Agr Univ Coll Forestry Jiangxi Key Lab Conservat Biol Nanchang Peoples R China

Globally, marine fish communities are being altered by climate change and human disturbances. We examined data on global marine fish communities to assess changes in community-weightedmean temperature affinity (i.e., mean temperatures within geographic ranges), maximum length, and trophic levels, which, respectively, represent the physiological, morphological, and trophic characteristics of marine fish communities. Then, we explored the influence of climate change and fishing on these characteristics because of their long-term role in shaping fish communities, especially their interactive effects. We employed spatial linear mixed models to investigate their impacts on community-weighted mean trait values and on abundance of different fish lengths and trophic groups. Globally, we observed an initial increasing trend in the temperature affinity of marine fish communities, whereas the weighted mean length and trophic levels of fish communities showed a declining trend. However, these shift trends were not significant, likely due to the large variation in midlatitude communities. Fishing pressure increased fish communities' temperature affinity in regions experiencing climate warming. Furthermore, climate warming was associated with an increase in weighted mean length and trophic levels of fish communities. Low climate baseline temperature appeared to mitigate the effect of climate warming on temperature affinity and trophic levels. The effect of climate warming on the relative abundance of different trophic classes and size classes both exhibited a nonlinear pattern. The small and relatively large fish species may benefit from climate warming, whereas the medium and largest size groups may be disadvantaged. Our results highlight the urgency of establishing stepping-stone marine protected areas to facilitate the migration of fishes to habitats in a warming ocean. Moreover, reducing human disturbance is crucial to mitigate rapid tropicalization, particularly in vulnerabl

关键词： climate warming community shifts fish length interactive effects marine tropicalization trophic levels

来源：评论

学校读者我要写书评

暂无评论

Enhancing automated health monitoring of road infrastructure through a hierarchical robust deep learning approach

引用

STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL 2024年

作者： Deng, fuwen Jin, Jiandong chen, Xiaobo An, Zhiyong Du, Yuchuan Shandong Technol & Business Univ Sch Comp Sci & Technol 191 Mid Binhai Rd Yantai 264005 Shandong Peoples R China Peking Univ Sch Elect Engn & Comp Sci Beijing Peoples R China Tongji Univ Coll Transportat Engn Shanghai Peoples R China

The life-cycle monitoring of road health conditions is crucial for precise and automated maintenance of road infrastructure. Despite this importance, existing machine vision-based methods for identifying pavement distresses are constrained in their susceptibleness to interference from factors such as shadows. To bridge this gap, the article introduces a novel hierarchical pavement structure monitoring model, which significantly enhances the task's performance. This proposed model integrates a de-shadowing module based on generative adversarial networks and an anchor-free instance segmentation model to provide pixel-level detection results, encompassing distress categories and morphological masks. A large-scale pavement image dataset was constructed, and a generative model was developed to expand pavement shadow samples. Experimental results show that the model achieved an average precision metric of 78.2%, surpassing existing baseline models. The effectiveness of the de-shadowing module was also validated through the significant improvements observed in various evaluation indicators. The proposed method demonstrates superior accuracy and robustness, holding promise for real-world road infrastructure monitoring systems.

关键词： Pavement health monitoring robustness enhancement hierarchical deep learning instance segmentation model

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：