基于深度学习的中文专利自动分类方法研究   总被引：2，自引：0，他引：2  

吕璐成  韩涛  周健  赵亚娟 《图书情报工作》2020,64(10):75-85

[目的/意义] 面向当前国内专利审查和专利情报分析工作中对于海量专利分类的客观需求，设计了7种基于深度学习的专利自动分类方法，对比各种方法的分类效果，从而助力专利分类效率和效果的提升。[方法/过程] 针对传统机器学习方法存在的缺陷，基于Word2Vec、CNN、RNN、Attention机制等深度学习技术，考虑专利文本语序特征、上下文特征以及分类关键特征，设计Word2Vec+TextCNN、Word2Vec+GRU、Word2Vec+BiGRU、Word2Vec+BiGRU+TextCNN等7种深度学习模型，以中国专利为例，选取IPC主分类号的"部"作为分类依据，对比这7种模型与3种传统分类模型在中文专利分类任务中的效果。[结果/结论] 实证研究效果显示，采用考虑语序特征、上下文特征及强化关键特征的深度学习方法进行中文专利分类具有更优的分类效果。  相似文献   

Game-Based Learning in Virtual Worlds: A Multiuser Online Game for Medical Undergraduate Radiology Education within Second Life     

Rocio Lorenzo-Alvarez  Teodoro Rudolphi-Solero  Miguel J. Ruiz-Gomez  Francisco Sendra-Portero 《Anatomical sciences education》2020,13(5):602-617

Game-based learning can have a positive impact on medical education, and virtual worlds have great potential for supporting immersive online games. It is necessary to reinforce current medical students' knowledge about radiological anatomy and radiological signs. To meet this need, the objectives of this study were: to design a competition-based game in the virtual world, Second Life and to analyze the students' perceptions of Second Life and the game, as well as to analyze the medium-term retention of knowledge and the potential impact on the final grades. Ninety out of 197 (45.6%) third-year medical students voluntarily participated in an online game based on self-guided presentations and multiple-choice tests over six 6-day stages. Participants and non-participants were invited to perform an evaluation questionnaire about the experience and a post-exposure knowledge test. Participants rated the experience with mean scores equal to or higher than 8.1 on a 10-point scale, highlighting the professor (9.5 ± 1.1; mean ± SD) and the virtual environment (8.9 ± 1.1). Participants had better results in the post-exposure test than non-participants (59.0 ± 13.5 versus 45.3 ± 11.5; P < 0.001) and a lower percentage of answers left blank (6.7 ± 8.4 versus 13.1 ± 12.9; P = 0.014). Competitive game-based learning within Second Life is an effective and well-accepted means of teaching core radiological anatomy and radiological signs content to medical students. The higher medium-term outcomes obtained by participants may indicate effective learning with the game. Additionally, valuable positive perceptions about the game, the educational contents, and the potential benefit for their education were discovered among non-participants.  相似文献   

Sociomaterial movements of students' engagement in a school's makerspace     

Kristiina Kumpulainen  Anu Kajamaa 《British journal of educational technology : journal of the Council for Educational Technology》2020,51(4):1292-1307

This study investigates the sociomaterial movements of student engagement in a school's makerspace. Here, we understand sociomaterial movements as emergent and relational, comprising complex dynamics of agency across students, teachers and materials in situated, culturally framed activities. Our study draws on data comprising 85 hours of video recordings of 9–12-year-old students' (N = 94) engagement in a technology-rich makerspace in a Finnish elementary school. The video data were transcribed and analyzed qualitatively using a multimodal interaction analysis. The sociomaterial movements were found to be displayed across a tension-laden continuum between (1) procedural activity—analysis and reflection; (2) individual activity—collaboration; (3) “doing school”—empowerment; and d) alienation—identification. Together, the study offers a potential approach for investigating and understanding the often overlooked workings of sociomateriality that constitutes students' emergent engagement and learning opportunities in science, technology, engineering and mathematics (STEAM) learning contexts.  相似文献   

高职院校学习预警平台的构建及其实证分析     

宋楚平  李少芹  蔡彬彬 《现代教育技术》2020,(5):94-100

开展信息化、智慧化预警工作是实现全面精准育人的重要途径,而预警应用平台无疑是预警工作的有力工具和实践载体,其设计方案应用效果的好坏直接决定着学习预警信息化愿景的达成度。文章在分析高职院校学习预警工作现状的基础上,从工作流程设计、平台框架设计和平台建设方法等3个方面提出了解决方案,并对平台的应用效果进行了实证分析,结果表明:平台的应用扩展了预警范围、提升了预警的精准度和时效性、提高了工作效率,且干预更为灵活、个性化,可为当前高职院校学习危机的实践改革提供新思路。  相似文献   

影响高职思政课学生学习效果的关键因素与对策     

高衍玉 《教育教学论坛》2020,(3):369-370

高职思想政治课的学习效果直接关系到立德树人这一根本任务的落实。在影响高职思想政治课学习效果的诸多因素中,学生自身需求是关键因素。针对这一关键因素,高职思想政治课教师需要综合运用多种措施满足学生需求,提高思想政治课学习效果,真正落实立德树人的根本任务。  相似文献   

A Hands-On Organ-Slicing Activity to Teach the Cross-Sectional Anatomy     

Lei Li  Lin Li  Yizhi Zuo 《Anatomical sciences education》2020,13(6):732-742

The presentation of pre-sliced specimens is a frequently used method in the laboratory teaching of cross-sectional anatomy. In the present study, a new teaching method based on a hands-on slicing activity was introduced into the teaching of brain, heart, and liver cross-sectional anatomy. A randomized, controlled trial was performed. A total of 182 third-year medical students were randomized into a control group taught with the prosection mode (pre-sliced organ viewing) and an experimental group taught with the dissection mode (hands-on organ slicing). These teaching methods were assessed by testing the students' knowledge of cross-sectional specimens and cross-sectional radiological images, and analyzing students' feedback. Using a specimen test on three organs (brain, heart, and liver), significant differences were observed in the mean scores of the control and experimental groups: for brain 59.6% (±14.2) vs. 70.1% (±15.5), (P < 0.001, Cohen's d = 0.17); for heart: 57.6% (±12.5) vs. 75.6% (±15.3), (P < 0.001, d = 0.30); and for liver: 60.4% (±14.5) vs. 81.7% (±14.2), (P < 0.001, d = 0.46). In a cross-sectional radiological image test, better performance was also found in the experimental group (P < 0.001). The mean scores of the control vs. experimental groups were as follows: for brain imaging 63.9% (±15.1) vs. 71.1% (±16.1); for heart imaging 64.7% (±14.5) vs. 75.2% (±15.5); and for liver imaging 61.1% (±15.5) vs. 81.2% (±14.6), respectively. The effect sizes (Cohen's d) were 0.05, 0.23, and 0.52, respectively. Students in the lower tertile benefited the most from the slicing experiences. Students' feedback was generally positive. Hands-on slicing activity can increase the effectiveness of anatomy teaching and increase students' ability to interpret radiological images.  相似文献   

The Potential Impact of COVID-19 on Student Learning and How Schools Can Respond     

Adam E. Wyse  Eric M. Stickney  David Butz  Amanda Beckler  Catherine N. Close 《Educational Measurement》2020,39(3):60-64

There is no denying the impact that the coronravirus disease (COVID-19) outbreak has had on many aspects of our lives. This article looks at the potential impact of COVID-19 on student learning as schools abruptly morphed into virtual learning environments using data from several instructional, practice, and assessment solutions offered by Renaissance. First, three hypothetical learning scenarios are considered using normative data from Star assessments to explore the potential impact on reading and math test performace. Next, data on Focus Skills are used to highlight which grades may have missed the most foundational math and reading content if instruction was stopped or reduced. Last, data from two of Renaissance's practice tools are used to evaluate whether students were practicing key skills following school closures. The article concludes that academic decline will likely occur but may be tempered by the increased use of practice tools; effects may look different for math and reading; and may impact grades and schools differently. As such, schools may need to leverage decision-making frameworks, such as the Multi-tiered Systems of Support/Response-to-Intervention (MTSS/RTI) framework, more than ever to identify needs and target instruction where it matters most when school begins in fall 2020.  相似文献   

基于负样本多通道优化SSD网络的钢铁厂提取     

卢凯旋  李国清  陈正超  昝露洋  李柏鹏  高建威 《中国科学院大学学报》2020,37(3):352-359

准确提取钢铁厂对去产能监测和环境保护具有重要意义。传统的人工目视解译方法效率低、成本高，无法满足开展大区域钢铁厂监测的需求。以深度学习目标检测网络SSD为基础，构建面向遥感影像钢铁厂提取的深度学习目标检测网络，提出maxout模块，将负样本通路优化为多分支结构，突出难分负样本特征并提升网络对无用特征的抵制效果。利用国产GF-1数据对京津冀地区的钢铁厂进行快速自动提取实验。与人工解译的钢铁厂点位数据的对比表明，该目标检测方法的提取精度达到80%以上。  相似文献   

动态雾计算网络中基于在线学习的任务卸载算法     

谭友钰  陈蕾  周明拓  王昆仑  杨旸  张武雄 《中国科学院大学学报》2020,37(5):688-698

任务卸载是雾计算的主要技术之一，即计算能力不足的节点将任务卸载给具有富余资源的节点帮助计算。以优化任务平均卸载时延和提升卸载服务成功率为目标，利用多臂老虎机理论为动态雾计算网络提出一种基于在线学习的任务卸载算法，可实时做出最优卸载决策。将该算法扩展到非稳定网络状态，使之可以动态追踪网络中节点的资源与环境变化，实时调整卸载决策。详细分析所提出算法的性能、复杂度和存储占用情况。仿真结果表明，这两种算法可达到的长期平均任务卸载时延均十分接近理想算法下的最优时延，卸载服务成功率也得到显著提升。此外，所提算法在非稳定的网络状态下能够追踪到计算资源与环境的变化。  相似文献   

Location extraction from tweets     

Thi Bich Ngoc Hoang  Josiane Mothe 《Information processing & management》2018,54(2):129-144

Five hundred million tweets are posted daily, making Twitter a major social media platform from which topical information on events can be extracted. These events are represented by three main dimensions: time, location and entity-related information. The focus of this paper is location, which is an essential dimension for geo-spatial applications, either when helping rescue operations during a disaster or when used for contextual recommendations. While the first type of application needs high recall, the second is more precision-oriented. This paper studies the recall/precision trade-off, combining different methods to extract locations. In the context of short posts, applying tools that have been developed for natural language is not sufficient given the nature of tweets which are generally too short to be linguistically correct. Also bearing in mind the high number of posts that need to be handled, we hypothesize that predicting whether a post contains a location or not could make the location extractors more focused and thus more effective. We introduce a model to predict whether a tweet contains a location or not and show that location prediction is a useful pre-processing step for location extraction. We define a number of new tweet features and we conduct an intensive evaluation. Our findings are that (1) combining existing location extraction tools is effective for precision-oriented or recall-oriented results, (2) enriching tweet representation is effective for predicting whether a tweet contains a location or not, (3) words appearing in a geography gazetteer and the occurrence of a preposition just before a proper noun are the two most important features for predicting the occurrence of a location in tweets, and (4) the accuracy of location extraction improves when it is possible to predict that there is a location in a tweet.  相似文献