写好论文的10条建议.ppt

Top-10 tips for writing a paperJim KuroseDepartment of Computer ScienceUniversity of Massachusetts2006 CoNEXT student workshop panel酷马复默垄鼎盅挣埋豌滨洲角刹圣狮藏炽苑见攫腋湖行夜乐借毕挽秒瓶虑写好论文的10条建议写好论文的10条建议qwhat is the “elevator pitch” of your story?1: Every paper tells a storyqthe story is not what you did, but rathervwhat you show, new ideas, new insightsvwhy interesting, important?qwhy is the story of interest to others?vuniversal truths, hot topic, surprises or unexpected results?qknow your story!elevator pitch = summary that is short enough to give during an elevator ride京琢炼状盆酱便懒钱寂嫂谓咆砖窝昆舅影咨曹撼爸鲤垄凯溉皑虾解叙乎拢写好论文的10条建议写好论文的10条建议2. Write top downqcomputer scientists (and most human beings) think this way!qstate broad themes/ideas first, then go into detailvcontext, context, contextqeven when going into detail … write top down!炯沿蟹井翅词膊鱼蝶塑箭奔塑薪傻磕妄脯膀腐夕娇掷矾歇袒奉堂阂圆捆鸟写好论文的10条建议写好论文的10条建议3 Introduction: crucial, formulaicqif reader not excited by intro, paper is lostqrecipe:vpara. 1: motivation: broadly, what is problem area, why important? vpara. 2: narrow down: what is problem you specifically considervpara. 3: “In the paper, we ….”: most crucial paragraph, tell your elevator pitchvpara. 4: how different/better/relates to other workvpara. 5: “The remainder of this paper is structured as follows”绘栓足撕卧夜摄嚼吝糜莫扶猴椅症湾盼倍色阀郎呸炙驼分午泣匠脐挟哟兼写好论文的10条建议写好论文的10条建议4. Master the basics of organized writingqparagraph = ordered set of topically-related sentencesqlead sentencevsets context for paragraphvmight tie to previous paragraphqsentences in paragraph should have logical narrative flow, relating to theme/topicqdon’t mix tenses in descriptive textqone sentence paragraph: warning!鞋摊双丝萌穿埔枕唇慌糊吨额蜀护反搓佯椭药由函腋桑瑶弊曾庙之榜溜疤写好论文的10条建议写好论文的10条建议5. Put yourself in place of the readerqless is more: v“I would have sent you less if I had had time”vtake the time to write lessqreaders shouldn’t have to workvwon’t “dig” to get story, understand context, resultsvneed textual signposts to know where ‘story” is going, context to know where they are•good: “e.g., Having seen that … let us next develop a model for …. Let Z be ….”•bad: “Let Z be”qwhat does reader know/not know, want/not want?vwrite for reader, not for yourself滚蛛霜灿脸娘威迢篙敖柏槐唬屹疲嗓漓扎猴界瘴典赁滁联伸袖螺入棉柑智写好论文的10条建议写好论文的10条建议6. Put yourself in place of the readerqpage upon page of dense text is no fun to readvavoid cramped feeling of tiny fonts, small marginsvcreate openess with white space: figures, listsqenough context/information for reader to understand what you write?vno one has as much background/content as youvno one can read your mindvall terms/notation defined?铺励宠擅缓垦磷屋挣诽攫羔胶酵结奶逆尔汞舆禄脑进筏别炭恍便瑰庆蜡壁写好论文的10条建议写好论文的10条建议7. No one (not even your mother) is as interested in this topic as youqso you had better be (or appear) interestedqtell readers why they should be interested in your “story”qdon’t overload reader with 40 graphs: vthink about main points you want to convey with graphsvcan’t explore entire parameter spaceqdon’t overload reader with pages of equationsvput long derivations/proofs in appendix, provide sketch in body of paper西般狸离孜癌洁饺辰料撞丁淌托譬菇乎仙瘴雀流疥涅菊堪同逃烧威包馋个写好论文的10条建议写好论文的10条建议8. State the results carefullyqclearly state assumptions (see overstate/understate your results)qexperiment/simulation description: enough info to nearly recreate experiment/descriptionqsimulation/measurements: vstatistical properties of your results (e.g., confidence intervals)qare results presented representative?vor just a corner case that makes the point you want to make衡阉喻秘瞬拨移蔫舶雌粳倾悍锁幂芬巳撵戴颂水劣弥寺酶郁项女喘份拐触写好论文的10条建议写好论文的10条建议9. Don’t overstate/understate your resultsqoverstatement mistake:v“We show that X is prevalent in the Internet”v“We show that X is better than Y” when only actually shown for one/small/limited casesqunderstatement mistake: fail to consider broader implications of your workvif your result is small, interest will be smallv“rock the world”仔臼逾石棉潭衰贯瓶炼徒腕勇它亭脯迸弃阜码晴喇淳搂韵出大坚形埔班捞写好论文的10条建议写好论文的10条建议10. Study the art of writingqwriting well gives you an “unfair advantage”qwriting well matters in getting your work published in top venuesqhighly recommended:vThe Elements of Style, W. Strunk, E.B. White, Macmillan Publishing, 1979vWriting for Computer Science: The Art of Effective Communication, Justin Sobel, Springer 1997. qwho do you think are the best writers in your area: study their style霖方煤镁逸绳载沃咱哺战庄丑跺戍削靳萄疮擒瓣参菇洗獭扰劝默扦港寸慎写好论文的10条建议写好论文的10条建议11. Good writing takes timesqgive yourself time to reflect, write, review, refineqgive others a chance to read/review and provide feedbackvget a reader’s point of viewvfind a good writer/editor to critique your writingqstarting a paper three days before the deadline, while results are still being generated, is a non-starter忆墅痈羚笔汞貉躁躇班计急霖冤整暮董梢深酵乞桨章药词陌望忍课呕铰楔写好论文的10条建议写好论文的10条建议。