Programmering - fra instruktion til modellering

Programmering - fra instruktion til modellering

Publication traditions and literature search Publication channels Journals, conference, workshops, etc. Peer-reviewing Bibliographic measures Citations, H-index and I-index Publication data bases Google Scholar DBPL computer science bibliography Semantic Scholar Bibliographic research indicator (BFI) Danish system to measure research production and distribute money to the universities AU Library in the Nygaard building (Rasmus T. Nielsen)

Publication traditions vary a lot from research area to research area If there is a conflict between the general information in this talk and the more specific information given by your advisor, you should always believe your advisor Publication channels Scientific papers can be published in different ways Journals are typically published 4-6 times a year (with 6-15 papers each time) Conferences are scientific meetings typically taking place each year (with 20-100 papers each time) there are also broad mega-conferences with hundreds of papers Workshops are informal scientific meetings, which typically covers an emerging area and have less strict quality demands for accepted papers Some conferences and workshops have tool sessions, where experimental computer tools are presented Some conferences and workshops have poster sessions, where the author stands in front of a large poster and explains it to anyone interested Within most sciences (naturvidenskab), journals are the only kind of publications which counts on your CV

In computer science things are different Here it is often more prestigious to publish at the best conferences than it is to publish in the best journals Many journals publish improved versions of the best papers from selected conferences (otherwise you cannot publish a paper twice) The maximal number of pages are typically larger for a journal paper than for a conference paper (hence it is possible to add more examples, all details of proofs, etc.) 2 Peer reviewing All (serious) journals and conferences are peer-reviewed (peer ligemand) This means that all submissions are read and evaluated by 3-5 referees (reviewers) who are experts within the area covered by the submitted paper The referees usually make a significant number of proposals for improvements and corrections of errors (including grammatical errors) Based on the referee reports it is decided which papers should be accepted/rejected

The best conferences and journals have acceptance rates which typically are 10-20% Hence, it is quite difficult to get papers accepted in these It is prestigious to act as referee for a good conference/journal It is something you list on your CV It also takes a lot of your time 3 Organization of conferences Each conference has a program committee with 20-30 PC-members lead by 1-2 two PC-chairs All PC-members are well-known researchers within the area covered by the conference (typically appointed by a steering committee in cooperation with the PC-chairs) The PC-chairs distribute the submitted papers to the PC-members, selecting those they believe to be the most knowledgeable within the area of the paper

PC-members may use sub-referees to help them evaluate a paper (this is often PostDocs or PhD students within the area) Based on the written referee reports PC-chairs propose whether the individual papers should be accepted, discussed or rejected Until a few years ago, it was typical to have a physical program committee meeting Today, nearly all conferences use electronic systems (special webboards) and virtual program meetings where people discuss and vote over a pre-determined period The accepted papers are collected in conference proceedings that are available at the conference (or shortly before) Authors are expected to improve their papers based on the comments in the referee report, but in practice there is seldom time to check this before the conference But if you don't fix serious errors, you will be in "bad standing" next time you submit a paper It will often be sent to the same referees, since they are the experts within your area 4

Organization of journals Each journal has an editorial board with 6-12 members lead by an editor-in-chief (sometimes assisted by a number of associate editors) All members of the editorial board are well-known researchers within the area covered by the journal (typically appointed by he editor-in-chief) When a paper is submitted, the editor-in-chief appoints a person to handle the submission (typically an associate editor og another member of the editorial board) The "handler" sends the paper to referees, who are experts within the area of the paper The referees may use sub-referees to help them evaluate a paper (this is often PostDocs or PhD students within the area) Based on the written referee reports, the "handler" proposes whether the paper should be accepted, rejected or can be resubmitted with minor/major improvements The final decision is made by the editor-in-chief The final revised papers are thoroughly checked by the original referees Good authors submit a document in which they describe how the different proposals

for improvements have been taken care of For proposals that have not been followed, detailed arguments are given For many papers, the final version of a paper is checked by a professional with linguistic background 5 Comparison of conferences and journals Each conference has a limited number of time slots available When all slots are filled, additional papers must be rejected although some of them may be of high-quality Such papers can be re-submitted next year or submitted to another conference It is wise to improve the papers before resubmission since it is likely that they will have one or more referees who have seen the old version The turn-around time is short (typically 4-6 months from submission to conference) There is an oral presentation at the conference (typically 15-20 minutes, followed by a short discussion)

Journals have "unlimited" space Accepted papers are queued and will be published when there is space in the journal If a paper is found good enough, it will be accepted The turn-around time is long (typically 1-2 years from submission to publication) Submissions can be done at any time There is no oral presentation of the paper Some journals have "special sections" covering a narrow area, e.g. with the best papers from a conference Authors are invited to publish, and papers will usually be accepted if they are of decent quality The space is limited (to some extend), and the turn-around time medium

6 Workshops and tool / poster sessions Workshops are similar to conferences but with less strict reviewing Papers with interesting ideas may be accepted even though they appear to have some flaws/gaps Some workshop are totally open without peer-reviewing Tool and poster sessions Rules and quality vary a lot from event to event It is a good way to show ongoing work and have discussions with colleagues, but they are seldom useful for your CV There are exceptions. As an example, it is very prestigious to be allowed to present new interaction ideas at some of the best HCI-conferences There is a significant number of bogus journals and conferences All (or most papers) are accepted with no peer-reviewing or with referees who are

unknown (internationally) in their field Conferences are typically held at "touristic" places like Hawaii or a cruise ship It will harm your CV to publish or referee at such places (since it indicates bad judgement) 7 Bibliographic measures Other sciences operate with "impact factors" which describe the quality of journals Since the majority of computer science publications are in conferences, impact factors have little importance/value within our field Instead, it is common to look at the following quantitative measures for a researcher Number of citations The number of (peer-reviewed) papers that have a citation (reference) to a paper written by the researcher

h-index The number of well-cited papers for the researcher To have an H-index of N you must have N papers which each have at least N citations i10-index The number of papers which have at least 10 citations each Good measure for the number of "serious" publications made by the researcher 8 Be careful when you compare two researchers The (scientific) age has a big influence Obviously, young researcher have much smaller numbers than old researchers When a paper has been written, it takes months before it is published, and then several years before it is read and cited by other researchers The M-index tries to handle this by dividing the H-index with the number of years since the first publication

The research area has a big influence In mathematics, it is common to publish at most one paper per year This obvious means that a researcher has few papers, and moreover there are few other papers that can make citations to them Even within computer science, there are big differences from research area to research area In some research areas, it is common that a young active researcher (e.g. a PhD student or a PostDoc) publish 10-20 papers a year In other areas this number is much lower Hence, it only makes sense to compare citations, H-index and I-index for researchers who are within the same (or similar) research area have the same (or nearly the same) scientific age 9 Google Scholar

One of the most commonly used data bases to investigate citations etc. within computer sicence Includes conference papers which are excluded by most other data bases Each researcher maintains a profile, which makes the counts more reliable, since missing papers can be added, and wrong papers (e.g. by authors with the same name) can be removed Google scholar overview Kurt Jensen The last column counts citations made within the last 5 years The papers that are cited may be much older My publication profile is unusual I have published relatively few papers

But my papers (and text books) have very high citation numbers I have only published 3 papers over the last 10 years (since I have been heavily involved in the leadership of the department) 10 Google Scholar (papers ordered by citations) 11 Google Scholar (papers ordered by year) 12 Individual papers You can click on the individual papers to

See where it is published, citation statistics, the beginning of the abstract, etc. download the paper (in most cases) To download, you often need to access the data bases of different publishing companies such as Springer, Elsevier, ACM, etc. AU has licenses to many of these To use them, you must be on the AU network (or use vpn) For help, ask your advisor or our library staff Many CS researchers have copies of their papers on their personal home pages (which Google Scholar has a reference to) 13 Search for papers without specifying the author About 29,400 results (0,05 sec)

Different search criteria by which you can limit the search In some research areas, it may also be important to search for patents 14 DBLP computer science bibliography Contains 5 million journal articles, conference papers, and other publications in computer science Covers all important journals and many conferences Maintained by the University of Trier, Germany since 1993 A lot of different information is available

15 Semantic Scholar Contains more than 40 million papers from computer science and biomedicine Covers all important journals and many conferences Maintained by the Allen Institute for Artificial Intelligence since 2015 Adds a layer of semantic analysis to the traditional methods of citation analysis A lot of different information is available 16 Bibliographic Research Indicator (BFI) List of journals and conferences (not individual papers)

Within each research area (such as computer science), all peer-reviewed journals and conferences are divided into two groups (by an appointed group of scientists) A publication in the best group gives 3 points, while a publication in the other group only gives 1 point Total number of points are used to distribute money between the Danish universities Conferences also count within computer science (but not in other areas) It can be used to compare universities, faculties and departments but only within Denmark 17 Time plan for the rest of this course Monday February 18, 11.15-13.00 How to write an academic paper Monday May 6, 11.15-13.00 How to make proper charts and graphs (by Hans-Jrg Schulz)

Saturday June 15 at 12 noon: Deadline for bachelor report Tuesday June 18, 11.15-13.00 How to make a good oral presentation Followed by a Poster presentation of all projects (from 13-15) June 24-28: Oral examination If you have proposals for additional lectures or other common activities), please send me a mail or make a posting on the webboard This also applies for common activities for the bachelor groups associated with a research group 18 That's all for now questions

19 AU Library, Katrinebjerg Rasmus T. Nielsen, Librarian, [email protected], 5335-170b www.library.au.dk Link www.library.au.dk What? Books, e-books, e-journals, reports etc. 20 AU Library, Katrinebjerg Seeking Information http://library.au.dk/en/students/information-seeking/ Link

Subject guides http://library.au.dk/en/subject-areas/electrical-andcomputer-engineering/ Link Avoid plagiarism http://library.au.dk/en/students/plagiarism/ Link 21

Recently Viewed Presentations

  • ftvs.cuni.cz

    ftvs.cuni.cz

    Publikační výsledky. J. imp. - původní nebo přehledový článek v odborném periodiku, který je obsažen v databázi Web of Science příznakem „Article", „Review", nebo „Letter", v roce 2013 i „ProceedingsPaper". J. SC. - původní nebo přehledový článek v odborném periodiku,...
  • State Criminal History Registry Checks

    State Criminal History Registry Checks

    To ensure the program is collecting, documenting, and reviewing necessary data; that it is on track to meet its stated objectives; and that it submits accurate and timely reports of AmeriCorps-related performance measure objectives.
  • Synonyms & Antonyms - WELCOME TO DERRY ENGLISH WITH MRS. HARR

    Synonyms & Antonyms - WELCOME TO DERRY ENGLISH WITH MRS. HARR

    Synonyms & Antonyms. Synonyms = Same. Antonyms = Opposite. Passage 1 - Gary & Lou. 1. Which is an antonym for the word unkempt as it is used in the passage? Passage 1 - Gary & Lou. 1. Which is...
  • Introduction to City Improvement Districts July 2019 Urban

    Introduction to City Improvement Districts July 2019 Urban

    Co-ordinate multi-disciplinal approach by including all City role players. Role of Observers (ToR) DO's: Must receive and review all materials circulated to directors . Attend and participate at meetings subject to ToR.
  • Elementary Linear Algebra Anton & Rorres, 9th Edition

    Elementary Linear Algebra Anton & Rorres, 9th Edition

    Determinants by Cofactor Expansion. Evaluating Determinants by Row Reduction. Properties of the Determinant Function. A Combinatorial Approach to Determinants. 2-1 Minor and Cofactor. Definition.
  • The Middle Ages - Mrs. Blair's World History Class

    The Middle Ages - Mrs. Blair's World History Class

    The Middle Ages Feudalism Feudalism Rises A system of landholding and governing It was based on an exchange of protection for other services lords Feudal System Lords give knights/vassals land (fief) in exchange for the knights' promise to defend the...
  • Basic Anatomy & Physiology - North Mason School District

    Basic Anatomy & Physiology - North Mason School District

    Basic Anatomy & Physiology Bones Provide: Protection Support Shape Primary Joints Hip Knee Shoulder Elbow Wrist Ankle Hip Femur Acetabulum Ligaments Labrum Muscles Hip Knee Femur Tibia/Fibula Patella Ligaments: MCL/LCL/PCL/ACL Meniscus Shoulder Humerus Glenoid Clavical Scapula Ligaments Muscles Labrum Elbow...
  • The ArrayList Class

    The ArrayList Class

    The ArrayList Class. An ArrayList is a complex data structure that allows you to add or remove objects from a list and it changes size automatically. The ArrayList class is part of the java.util package of the Java standard class...