Springer Nature培训2018年11月 - 广东医科大学.pptVIP

1700余种 76 1700余种 77 78 79 80 81 82 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 翻拍自河出书房新社一九九三年的《新文艺读本：寺山修司》 But please don’t forget, before internet, or even in today, this might be the way that inspiring great works. 73 74 75 12 13 14 15 16 17 18 19 20 21 22 23 24 在应用于推文的情感分析和情绪检测领域中，与构建自动分类器有关的主要问题之一是缺乏合适的训练集。考虑到手动注释训练集的繁琐以及直接从社交网络收集的数据中存在的噪声，本文中我们提出了一种迭代学习方法，它将远程监督与数据集修剪技术相结合。特别是，遵循“吃自己的狗食”的想法，我们已经将从不同的Twitter渠道获得的原始数据进行训练的分类器应用于相同的原始数据集，以便自动移除大多数情况。基于鹦鹉六种基本情绪模型，这种方法已经被用于获得更加优化的情感分类训练集。在取得的成果的基础上，我们认为训练集的自动过滤可以使远程监控方法在许多用例中更加有效。 25 26 27 28 29 30 31 32 33 34 35 36 37 Simpel, kan nog simpeler, blijft model; werkelijk onderzoek nooit sequentieel 38 39 40 41 42 43 Brevity (简短) Be clear (清晰) Be clear (清晰) Too much information! Difficult to pull the main claim of the paper out from the jumble of information provided. We need to be able to glance at the figures and understand them The axes labels of this graph can’t be understood without referring to the text Trend lines: add more information to an already busy graphic Reference to a previous graphic (‘Vostock’ and ‘GRIP’) Be clear (清晰) Tip 4 Decide early on where to publish 提前决定投哪本期刊 This will help shape your study, based on the goals needed for publication in your target journal. Will help define the form of study and advance required. 针对期刊对文章的要求进行研究，有助于把握研究方向和创新性。 Look at journal’s aims and scopes page 仔细阅读该期刊所涵盖领域及对文章的要求 Think about how you will structure your papers when you design your experimen 在设计实验时就开始思考文章架构 What controls and statistical tests are needed? 设置哪些对照组，使用何种统计方法 What collaborators / co authors should you work with to complete study? 需要和哪些共同作者合作才能完成该研究 What is your aim with study? What are you trying to show / prove? 研究目的是什么？想要表现或证明什么？ Tip 5 Quality is everything 质量决定一切 Try to publish in as high a quality journal as you can. 尽可能发表在质量最高的期刊上 One great study is better than several lesser quality ones 一篇高质量的文章多篇内容相似的一般文章 Avoid trying to publish lo