abezerianos - career seminar-lecture 4 slidesanab/teaching/careerseminar/lecture4.pdf · final...

Research Seminar

session 4 Anastasia Bezerianos

Wendy MackayTAs: Stact Hsueh and Eugenie Brasier

contact: [email protected]

https://www.lri.fr/~anab/teaching/CareerSeminar/

1 Career Seminar-Lecture 4 slides - 15 October 2020

Homework Assignments

#3 Review a paper


Reminder Assignment #2: research notebook

Due: today

Create your personal research notebookChoose paper, electronic or hybrid

Continue for the rest of the semester …Keep track of what you readSketch and record ideasDATE every entryAdd KEYWORDS to every entry

Upload a new entry every week (see assignment 2)


Staying organized

If you continue a career in research–or anything else– you will write many reviews, papers, etc.

Give files useful names!Include dates, titles, authors

Organize files according to project:Not: projectdescription.pdfBut: 2018.Octopocus.Experiment.1.doc


What happens to a review?

Program committeeReviewer 2

Reviewer 1

Author

Subcommittee chair

Reviewer 3

paper review meta review rebuttal

Associate chair (primary)

Associate chair (secondary)


Assignment #3: Write a paper review

Due: 15th October

1. Reread the paper from Assignment #1or better : choose a new article

2. Read the paper and take notes in your notebook3. Fill out the review form

https://www.lri.fr/~anab/teaching/CareerSeminar/ 4. Upload to the drive we indicated


Assignment #3: Review a paper

Review the paper as if you are an external reviewer

This means:You DO know where it was submittedYou DO NOT know who the authors areYou DO NOT know the impact the paper will have

You must summarize the paperBUT ALSO give your opinion


Review questions

Rank the paper:[ ] Overall rating (1=poor, 5=excellent)[ ] Reviewer expertise (1= no knowledge, 5=expert)

Please answer the following questions:Summarize the contribution to the field (1-2 sentences)Provide a detailed review (2-5 paragraphs)Indicate specific suggestions for improvement


What is a good review?

Your first audience is the program committeeEither to support the author (accept)or argue against (reject)

Short, vague reviews are USELESS and will be ignored

Avoid middle ratings (3 out of 5) Program committees prefer positive or negative ratings, not neutral ones.


Reviewing the reviewer

Editors or Meta-Reviewers also review their reviewers:Do they re-explain the main point(s) of the paper?Are the criticisms specific?Do they contribute additional references if that is a complaint?Do they articulate the contribution of the paper and related it to

past work?Do they state what the value of the contribution is to an

attendee of the conference or reader of the proceedings?Do they have some clear ideas about how the paper can be

improved or extended to increase its value?


Assignment #3: Review form

1. Rank the paper:[ ] Overall rating (1=poor, 5=excellent)[ ] Reviewer expertise (1= no knowledge, 5=expert)

2. Summarize the contribution to the field (1-2 sentences)3. Provide a detailed review (2-5 paragraphs)4. Indicate specific suggestions for improvement5. Include the ACM-style paper reference:

Eric A. Bier, Maureen C. Stone, Ken Pier, William Buxton, and Tony D. DeRose. 1993. Toolglass and magic lenses: the see-through interface. In Proceedings of the 20th annual conference on Computer graphics and interactive techniques (SIGGRAPH '93). ACM, New York, NY, USA, 73-80.


Class Exercise: Writing Walkthrough


Exercise: Writing Walkthrough

Structured Walkthroughs (Yourdon, 1979)

Goal: Find bugs in codeTechnique: Systematic step-by-step analysis

of a document by a small group

Principles: Line-by-line analysisConstructive criticismLimited time


Writing Walkthrough: Roles

Author explains:Document state: early draft, almost done?Publication: audience? deadlines?Criticism level: structure? style? grammar?

Moderator manages session:Timing: 5 min. to read, 15 min. to analyzeKeep comments constructive, avoid debates

Participants (including author!):Read through document onceWrite comments on paperOffer constructive comments out loud


Writing Walkthrough: Procedure

Create a group of authors:4 people, 20 minutes each = 1 hour 20 minutes

Preparation:Copy selected document parts (max. 1 page)

Procedure per author05 min: Everyone reads and annotes text15 min: Start with sentence one:

proceed line by line: identify problems


Writing Walkthrough: Rules

Constructive criticism:Be positiveGrammatical errorsLogic errors“I did not understand this”

Do not debate: it wastes time!Participants identify problems

and suggest solutionsAuthors can accept solutions

… or not!


Remember

DO NOT discuss- make the point and move on

The author decides … later

Your goal to go get through as many lines as possiblenot to ‘win’ an argument about a point


How do you choose an internship

or “stage”?


First, decide what you want to do next

Do you want to teach?- University professor track

Do you want to do research?- Academic research- Industry research

Do you want to go into industry?- large corporation- small company- create a startup

Something else?


Where do you want to work?

In your own country?other country?

In France? Europe? North America?Asia? Other parts of the world?


Internships

Introduce you to real-world research lab conditions- but each lab is different

Help you decide what you’d like to do next

If you do well, in a good research lab,provide you with a critical start on a research career

If you do well, in a good industry lab,provide you with a critical start or even position in industry

BUT not written in stone …


Evaluate the company you are going to

Big company? Small company? Start-up?

How do you know if the group is any good?


Evaluate the company you are going to

Big company? Small company? Start-up?

How do you know if the group is any good? ProfitProductPeople working in itColleaguesOthers ?


Key outcomes

Conduct impactful, applied workFigure out if you really like working in a company

Get a great letter of recommendation for industry

Start your professional network

Apply for a job with them


Evaluate the lab you’re going to

Make sure you are in a good lab!Poor training now can hurt your career

How do you know what’s a good lab?Publications & Bibliometrics (later on …)

How do you know if your advisor is any good?Publications & BibliometricsPublications with students

Differences between senior and junior researchers


Key outcomes

Conduct publishable research if interested in research(ideally, publish it!)

Figure out if you really like doing researchand whether or not you are good at it

Get a great letter of recommendation

Start your research network

Apply for a Ph.D. grant


Internships (research or industry)

Choosing a topic

Easiest: Find a lab, pick one of their topics

Best: Suggest something related to their work, but also related to what you want to do(research labs can be more flexible to diff topics)

… How to find a lab?


How do you choose a Ph.D.?


What is a Ph.D. (Doctor of Philosophy)?

In 3+ years you will:become a world expert on your topic, andwrite a doctoral dissertation: 150-300 pages of original research

You must develop a thesis: something that you believe!State it clearlyJustify it with your researchMake it the foundation for future research


Levels of education

High school 3-4 yearsUniversity 3-4 years under-graduate

Masters 1-2 years graduate

Ph.D. 3-4 years (France)4-8 years (US, UK, Germany…)

Post-doc 1-2 years

Masters degrees are pretty similar in Europe and North AmericaLength and scope of Ph.D.s can vary greatly


Advanced research careers

Government researchJunior research scientist Principle research scientistChargé de recherche => Directeur de recherche

University researchAssistant professor ProfessorMaître de conférence => Professeur

US Start with a temporary position~5-7 years: try for tenure (a permanent position)

or else get fired

France Start with a permanent position (almost)~7 years: write an habilitation,

apply for senior position


Other jobs that benefit from a Ph.D.

IndustryStartupsSmall companiesLarge companies

Industry researchMicrosoft, Google, Apple

AcademiaGovernment research (Inria, CNRS, National labs, Max Planck)University position


How do you find a Ph.D. grant?

You must find aTopic usually related to your masters, not alwaysLab at or associated with a universityAdvisor talk to them!

Sources of Ph.D. funding:

Government grantsLocal: UPSud / Paris-Saclay Ecole Doctorale bourseNational: Inria CORDI grant, Digicosme, Labex …

Thesis advisor’s research grantsGovernment (Country grands like ANR, EU, NSF, NSERC)Industry (CIFRE, Inria-MSR)


Different universities–different processes

Each country has a unique processIndividual universities differ as well

You will have to investigate:application proceduresdeadlines (may not fit your internship/master studies)entry requirementsfunding opportunitieslength of the degreecontent of the degreejobs when you are done


Entry requirements

Curriculum vitae (C.V.)includes your education and work history

Transcripts (grades)translate (not only the language, but what they mean)

TestsT.O.E.F.L. (Test of English as a Foreign Language)G.R.E. (Graduate Record Exam, US/Canada)

Letter of motivationmake it specific to the position

Letters of Recommendationchoose people who know you, not just ‘famous’ peoplepeople (usually) honest in their level of support


For Ph.D. students (but also internships) …

What is the problem?How you frame the problem influences the solution

What is your solution?Framing the problem well often points to a solution

Why is this problem difficult?Why haven’t others solved it already?

What is your evidence?Empirical finding Technical advanceMathematical algorithm Theoretical foundationReal-world impact Something else?


Questions to answer

What is your (research) community?Who has the problem? Who needs the solution?

How does your work compare to existing work?Who else is addressing the problem?Why are they different? (think both research and industry)

Who will benefit from your work?Users? Researchers? Industry?

What is the potential impact of your work?Short-term? Long-term?


State your thesis

One sentenceWhat do you believe, based on your work?

One paragraphAdd the problem definition and context

One pageAdd the justification: empirical, theoretical, technical

How will others cite your thesis?


Can you describe your thesis?

1. Elevator One-two sentences2. Cocktail party 3-4 paragraphs3. Short talk 10 minutes4. Long talk 30 minutes5. Defense 60 minutes

In each case, you need to communicatethe key point


Letters of Motivation


Letter of motivation – What is it?

Also: Statement of purposePersonal essay

Common for Ph.D., scholarship and job applications

An essay that explains what you hope to achievein the program you are applying to

as well as what qualifies you to be the top candidatefor the position

Shows your personality, not just factsBoth what you can contribute and how you will benefitDon’t be vague, use specific examples


Letter of motivation – How to start

Look up examples: BUT DO NOT COPY THEM!Write a unique letter for each application

Figure out the exact format to use

Avoid restating your CV

Draft a set of answers to key questions:Why this program/job/grant? Why you?

Write, get feedback, rewrite …


Letter of motivation – Why?

Take the perspective of the person who will read it… and letters by perhaps 100 others

Are you what they are looking for and how can they tell?

Do you come across as: honest?capable of clear thinking?highly motivated?

Don’t be too modest … or too arrogant

You should be answering “why”


For examplehttp://ic.daad.de/imperia/md/content/islamabad/letter_of_motivation.pdf


Optional exercise: Write a letter of motivation

Work in groups of twoFind a (research) lab where you would like to work

Write a two-paragraph letter of motivation:Why do you want to work in this lab and with whom?

(Ph.D.? internship? research project?)Why are you a good candidate?

(relevant skills, experience, degrees)What would you like to do?

(specify a research question and how it relatesto their work)


Letters of recommendation


Letters of recommendation

Researchers: Write articles - Review articles Write letters - Review letters

Any position or grant that you apply for

will require one or more letters of recommendation

Take the perspective of the: letter writer : to encourage them to write a good letter letter reader: to encourage them to hire you or give you the grant


What a good letter looks like

It is my pleasure to recommend Ann SMITH for a Ph.D. grant at the Université Paris-Saclay. Ann was one of the top students in my XXX class last year and I supervised her Master’s thesis project, “XXX”, from March-August 2015.

Ann is highly motivated and took advantage of her strong programming skills to build a novel interface that <what it does>. She has an excellent level of written and spoken English and I was impressed by her creativity in solving complex problems. Ann is friendly and quickly became an active member of the lab.

I believe Ann will benefit greatly from the opportunity to work in your lab, and she has the potential to become an excellent researcher.

I thus recommend Ann with no reservations. Please feel free to contact me If you have any further questions,


What a bad letter looks like

Ann Smith has asked me to write a letter of recommendation for a job as a programmer in your organization. Ann was a student in my Introduction to Programming class and received a good grade. She completed all the assignments and, although she is very quiet, I believe she learned the material.

I have not worked with her directly, but she is working with a colleague on an interesting project for her Master’s thesis. However it is too early to know how that project turns out.

I would like to recommend Ann for this position. Please feel free to contact me if you have any further questions,


What is the difference?

One is specific, warm and clearly positive. The other is vague, with very little information


Letters of Recommendation

“Damned with faint praise”

Poor letters are short, vague and generalJean attended my class and performed well.

Good letters are longer, detailed and personal.Jean was a top student in my Research Seminar. He always came to class well prepared and contributed insightful comments during class discussions. I was particularly impressed by the review he wrote on Smith’s classic paper, in which he ….


Asking for a Letter of Recommendation

Decide how to contact the letter writerIn person? Email? Paper letter?

ASK first if they are willing to give you a (strong) letterExplain exactly what they need to do

Don’t wait until the last minute (a month in advance is good)Give them a real deadline and double-check (remind them)



Provide all relevant documentsDon’t expect them to stand by the printer and print your documentsor look over their notes for your marks or projects

Make it easy!Add info of how they know you, how you did in their class, …Add a post-it note to show where to sign (if paper)Addressed, stamped envelope or a website to upload (with appropriate passwords)or an email address to send it to


Ask for a letter of recommendation

Letters of recommendationAdministrative documents

You are asking an extremely busy person to do you a personal favorthat probably does not help that person

Put yourself in their shoes and make it EASY for them



Give them all the relevant information

-or-

Either write it yourself and let them edit it(This is really hard! But some people ask for it.)



Explain why you need the letterApplying for a job, Ph.D., grant

Remind them how they know youWhen did you first meet?What have you done together?

Accomplishments they should mention

Personal characteristics to emphasize



How do you get good letters of recommendation?

Help the instructor remember you …

be memorable! (in a good way)

remind them of what you did

volunteer



Confidentiality

Will you be able to read the letter …

or not?

Letters that you can read are less valuable than anonymous letters


Some things to consider

Keep your request short Add extra information after your signature

Assume the recommender is on your sideGive them the information needed to make your case

Even if you know the person wellGive them detailed information: do not make them search

Never ask: “Do you remember me?Instead, remind them of what you did together

Some people / cultures are more formalIf you are not sure, formal is safer


Homework Assignment #4:

Ask for a

Letter of recommendation



Part 1: Write an email request for a letter of recommendation“Would you be willing to write a strong letter …”

Who are you? How do they know you?Why are you qualified? What is the letter for?When is it due? What is the process?

Part II: Write a draft letter for them to editSubject: Recommendation for <Who?>, for a <What??

Dear <Name>:


ExampleTo: [email protected]: Letter of Recommendation for Ann SMITH

Dear Professor Mackay:

I hope your semester is going well <something personal, reminder of how you know the professor>.

I have a favor to ask. I have decided to apply to <research program/job/etc> and I hope that you can write me a strong letter of recommendation. <Reminder of what you have done that merits a letter.>

The letter must be <sent to WHO? Uploaded where?> and deadline is <specific date>. I have attached my C.V. <and a quick summary of my more recent work>.

Please let me know if you will be able to write the letter and let me know if you have any questions. Thank you very much.

Best regards, Ann Smith


Assignment #4

Assignment #4: Ask for a letter of recommendationDue: 22th of Oct, upload to

https://drive.google.com/drive/folders/1CndvCwW4HqdHM_fioW5MQ7GXobydq9Rg?usp=sharing

name your file LASTNAME_Firstname-A4-Recommendation.pdf

Optional: Assignment #3b: Review a research paperBased on comments from last week’s exerciseyou may revise it to improve the reviewand turn in a version to be graded on 22th October

(upload in the location for Assignment 3, new file name should beLASTNAME_Firstname-A3b-Review.pdf


abezerianos - career seminar-lecture 4 slidesanab/teaching/careerseminar/lecture4.pdf · final...

Documents