How to Write a Scientific Paper Prof. Bruno Castro da Silva - PowerPoint PPT Presentation

How to Write a Scientific Paper Prof. Bruno Castro da Silva Institute of Informatics - UFRGS Heavily based on http://www.meneguzzi.eu/felipe/presentations/paper-writing.pdf

Report for a Class Project vs. Scientific paper

Report for a Class Project Scientific paper

Report for a Class Project Scientific paper

• Science is about being skeptical • when trying to understand something about the world 
 • You don't need to believe in other scientists • It’s their duty to convince you • And you should be able to reproduce their findings 
 • Scientific papers are the way scientists do this

• Science is about being skeptical • when trying to understand something about the world 
 • You don't need to believe in other scientists • It’s their duty to convince you • And you should be able to reproduce their findings 
 • Scientific papers are the way scientists do this

• Science is about being skeptical • when trying to understand something about the world 
 • You don't need to believe in other scientists • It’s their duty to convince you • And you should be able to reproduce their findings 
 • Scientific papers are the way scientists do this

A Scientific Paper

A Scientific Paper

A Scientific Paper • Why? To propose an idea to solve a problem

A Scientific Paper • Why? To propose an idea to solve a problem • How? You need to convince other people • (1) That it’s worth it (the problem is relevant to someone) 
 • (2) That the problem has not yet been solved • or that you can solve it better than other people 
 • (3) And that you really did (at least partially) solved it • or that you found a better way of solving it

A Scientific Paper • Why? To propose an idea to solve a problem • How? You need to convince other people • (1) That it’s worth it (the problem is relevant to someone) 
 • (2) That the problem has not yet been solved • or that you can solve it better than other people 
 • (3) And that you really did (at least partially) solved it • or that you found a better way of solving it


 A Scientific Paper • Why? To propose an idea to solve a problem • How? You need to convince other people • (1) That it’s worth it (the problem is relevant to someone) 
 • (2) That the problem has not yet been solved Jiwon Han (…) received this prize for studying the movement of coffee when a person • or that you can solve it better than other people 
 carries it while walking backwards 
 Han demonstrated his findings – that we are more likely to spill coffee when walking • (3) And that you really did (at least partially) solved it backwards with it – on stage during the ceremony, holding a disposable cup of coffee • or that you found a better way of solving it (Ig Nobel Prizes 2017)

A Scientific Paper • Why? To propose an idea to solve a problem • How? You need to convince other people • (1) That it’s worth it (the problem is relevant to someone) 
 • (2) That the problem has not yet been solved • or that you can solve it better than other people 
 • (3) And that you really did (at least partially) solved it • or that you found a better way of solving it

A Scientific Paper • Why? To propose an idea to solve a problem • How? You need to convince other people • (1) That it’s worth it (the problem is relevant to someone) 
 • (2) That the problem has not yet been solved • or that you can solve it better than other people 
 • (3) And that you really did (at least partially) solved it • or that you found a better way of solving it

A Scientific Paper • Why? To propose an idea to solve a problem • How? You need to convince other people • (1) That it’s worth it (the problem is relevant to someone) 
 • (2) That the problem has not yet been solved • or that you can solve it better than other people 
 • (3) And that you really did (at least partially) solved it • or that you found a better way of solving it

Structure of a Scientific Paper • Title • Abstract • Introduction • The problem • My idea • The details • Related work • Conclusions and further work

Structure of a Scientific Paper • Title (1000 readers) • Abstract (4 sentences, 100 readers) • Introduction (1 page, 100 readers) • The problem (1 page, 10 readers) • My idea (2 pages, 10 readers) • The details (5 pages, 3 readers) • Related work (1-2 pages, 10 readers) • Conclusions and further work (0.5 pages)

Structure of a Scientific Paper • Title (1000 readers) • Abstract (4 sentences, 100 readers) (1,2,3) • Introduction (1 page, 100 readers) (1,2,3) • The problem (1 page, 10 readers) (1,2) • My idea (2 pages, 10 readers) (3) • The details (5 pages, 3 readers) (3) • Related work (1-2 pages, 10 readers) (2) • Conclusions and further work (0.5 pages) (1,2,3) (1) - What is the problem + why it’s worth it 
 (2) - Why it hasn’t been completely solved yet 
 (3) - How I did solve it (partially?)

A Scientific Paper

A Scientific Paper Fallacy 
 You need to have a fantastic idea 
 before you can write a paper

A Scientific Paper Fallacy 
 you You need to have a fantastic idea 
 before you can write a paper

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper Ultimate goal is to convey an idea

Structure of a Scientific Paper • Title (1000 readers) • Abstract (4 sentences, 100 readers) (1,2,3) • Introduction (1 page, 100 readers) (1,2,3) • The problem (1 page, 10 readers) (1,2) • My idea (2 pages, 10 readers) (3) • The details (5 pages, 3 readers) (3) • Related work (1-2 pages, 10 readers) (2) • Conclusions and further work (0.5 pages) (1,2,3) (1) - What is the problem + why it’s worth it 
 (2) - Why it hasn’t been completely solved yet 
 (3) - How I did solve it (partially?)

Structure of a Scientific Paper • Title (1000 readers) • Abstract (4 sentences, 100 readers) (1,2,3) • Introduction (1 page, 100 readers) (1,2,3) • The problem (1 page, 10 readers) (1,2) • My idea (2 pages, 10 readers) (3) • The details (5 pages, 3 readers) (3) • Related work (1-2 pages, 10 readers) (2) • Conclusions and further work (0.5 pages) (1,2,3) Learning Unknown Markov Decision Processes: A Thompson Sampling Approach The Expxorcist: Nonparametric Graphical Models Via Conditional Exponential Densities

Structure of a Scientific Paper • Title (1000 readers) • Abstract (4 sentences, 100 readers) (1,2,3) • Introduction (1 page, 100 readers) (1,2,3) • The problem (1 page, 10 readers) (1,2) • My idea (2 pages, 10 readers) (3) • The details (5 pages, 3 readers) (3) • Related work (1-2 pages, 10 readers) (2) • Conclusions and further work (0.5 pages) (1,2,3) (1) - What is the problem + why it’s worth it 
 (2) - Why it hasn’t been completely solved yet 
 (3) - How I did solve it (partially?)

Abstract • Used by peers to decide which papers to read 
 • Four sentences: • State the problem • Say why it’s an interesting problem • Say what your solution achieves • Say what follows from your solution

Abstract • Used by peers to decide which papers to read 
 • Four sentences: • State the problem • Say why it’s an interesting problem • Say what your solution achieves • Say what follows from your solution

Abstract • Used by peers to decide which papers to read 
 • Four sentences: • State the problem • Say why it’s an interesting problem • Say what your solution achieves • Say what follows from your solution • Example 
 • Robots cannot fly • This is a pity, because flying robots would be useful to deliver mail • We propose a method by which humanoid robots can fly at subsonic speeds by using 
 a magic engine • We compare our method with two sample state-of-the-art techniques and show that 
 ours perform better when there is no wind. We expect our method will allow for faster 
 mail delivery

Structure of a Scientific Paper • Title (1000 readers) • Abstract (4 sentences, 100 readers) (1,2,3) • Introduction (1 page, 100 readers) (1,2,3) • The problem (1 page, 10 readers) (1,2) • My idea (2 pages, 10 readers) (3) • The details (5 pages, 3 readers) (3) • Related work (1-2 pages, 10 readers) (2) • Conclusions and further work (0.5 pages) (1,2,3) (1) - What is the problem + why it’s worth it 
 (2) - Why it hasn’t been completely solved yet 
 (3) - How I did solve it (partially?)

Introduction • Describe the problem 
 • State your contributions 
 • Restate key items from the abstract • (motivation/why the problem is interesting, etc) 
 • And use intuitive language