HOW TO GET A PH.D.: Methods and Practical Hints ( ) /to_phd_2006.

Transcription

1 HOW TO GET A PH.D.: Methods and Practical Hints ( ) /to_phd_2006.html Aarne Mämmelä HOW TO GET A PH.D. Methods and Practical Hints Dr. AARNE MÄMMELÄ Research Professor (VTT), Docent (Univ. Oulu, HUT) VTT, R&D, Telecommunications Kaitoväylä 1, P.O. Box 1100, FIN Oulu, Finland aarne.mammela@vtt.fi Tel (VTT), (direct), (direct) Fax Aarne Mämmelä 2 1

2 Contents of the Course Lectures on Tuesdays at (no lecture on ) Exam (Part I, 3 credits), additional exams if needed compulsory lectures and the textbook Part I must be accepted before Part II is accepted Textbook (several books ordered to the library) P. Bock, Getting It Right: R&D Methods for Science and Engineering, Academic Press, Course work (Part II, 5 credits, voluntary addition) proposal for requirements feedback final report Grades: Aarne Mämmelä 3 Course Work Literature search, and analysis, and reporting of the findings Define a research problem, suitable keywords, and a starting point (a recent original paper) for your search Use at least three databases: electronic library (IEEE Xplore, ACM Digital Library, SPIE Digital Library, etc.), and abstracts database (INSPEC, COMPENDEX, etc.) citation index (SCI-Expanded, SCOPUS, etc.) Final report about pages More detailed instructions on the course page Aarne Mämmelä 4 2

3 PROGRAM (1) - changes possible I Session at 2-5 pm (PR119) 1. Aarne Mämmelä, Literature Reviews: Existing Knowledge from Data Bases 2. Aarne Mämmelä, Final Result: a Scientific Publication II Session at 2-5 pm (PR119) 3. Lisa Lena Opas-Hänninen, Writing in English III Session at 2-5 pm (PR119) 4. Aarne Mämmelä, Research Methods: From Problem and Hypothesis to Experiments IV Session at 2-5 pm (PR119) 5. Tauno Jokinen, Research Methods in Industrial Engineering and Management 6. Olli Silven, Peer Review Process: the Task of a Referee V Session at 2-5 pm (PR119) 7. Lauri Pirttiaho, Being a Ph.D. in a Big Company Aarne Mämmelä 5 PROGRAM (2) - changes possible VI Session at 2-5 pm (TF104) 8. Erkki Oja, Experiences of a Senior Researcher 9. Riitta Keiski, Research Methods in Process and Environmental Engineering VII Session at 2-5 pm (IT138) 10. Maija Kallinen, History of Science and Technology 11. Riitta Keiski, Research Ethics VIII Session at 2-5 pm (IT138) 12. Helka-Liisa Hentilä, Research Methods in Architecture 13. Kari Leppälä, How to Do Scientific Research on Engineering and Technology IX Session at 2-5 pm (PR119) Program to be confirmed. X Session at 2-5 pm (PR119) 14. Petri Ylikoski, Philosophy of Science Aarne Mämmelä 6 3

4 Additional Material: Vocabulary abduction (abduktio) inference to the best explanation [Honderich05], opposite to deduction, facts are explained by a theory which best explains the facts, sometimes assumed to be a form of induction abrubt change (äkillinen muutos, hyppy) jump, usually not amenable to analysis, opposite to continuity [Wolfram02], see anomaly, catastrophe theory absorption cost (omakustannusarvo, OKA) prime cost abstract (abstrakti) opposite to concrete abstract (abstrakti, tiivistelmä) a one-paragraph summary of a document after the title and list of authors, an abstract can be informative (preferred) or descriptive [Young02], includes what the author has done, how it was done (if it is important), principal results (numerically, when possible), and significance of the results [Spectrum65], see abstracts database, executive summary, extended abstract, introduction, conclusions Aarne Mämmelä 7 Additional Material: Bibliography 4.7 ENGLISH-FINNISH DICTIONARIES OF ENGINEERING TEPA term bank, (This term bank is maintained by the Finnish Terminology Centre, Sanastokeskus TSK, and includes vocabularies from SFS standards and many other vocabularies, altogether over terms. The newest suggested terms for information engineering are in Termitalkoot where Julkaistut suositukset cover the published recommendations.) J. K. Talvitie et al., Englanti-suomi tekniikan ja kaupan sanakirja [English- Finnish dictionary of technical and commercial terms]. 10th ed. Tietoteos ky, (More that words. A similar electronic dictionary is MOT WSOY Enteka 3.0a available from Available at in NetMOT sanakirjasto, but only in the university campus network.) Aarne Mämmelä 8 4

5 Simplified Research Process Idea Literature review Note. In practice the process is highly iterative. Problem and hypotheses Experiments/ analysis System (prototype) Theory/paper (new knowledge) Aarne Mämmelä 9 Abstract of My Three Lectures The students will be introduced to systematic research methods in engineering. Research is a demanding learning exercise. I will proceed from bottom up, which is the by far the most natural way to learn and teach. During the first lecture I will present the most efficient ways to find literature using bibliographies and databases. During the second lecture I will explain how to organize a scientific publication according to a top down structure, which is the best way to present scientific results. The hardest part is to understand the research methods themselves. In research a lot of creativity is needed. I present a method that I call break down and reassemble, referring to analysis and synthesis, and it will encourage the researcher towards the bottom up approach, many times based on analogies. This will be the topic of the third lecture. Concrete examples will be given as much as possible, usually before the corresponding theoretical summary. Most of the examples are from electrical engineering, but the methods can be used in any empirical research Aarne Mämmelä 10 5

6 I LITERATURE REVIEWS: Existing knowledge from data bases Introduction Bottom-up learning process Taxonomy of publications Example data bases Taxonomy of databases Publishers How to find existing knowledge Outlining a literature review Conclusions Abbreviations References Aarne Mämmelä 11 Table of Contents Introduction 13 Learning process 16 Taxonomy of publications (1) 17 Example data bases 23 Amazon 24 IEEE Xplore (IEEE/IET Electronic Library, IEL) 33 INSPEC 38 Science Citation Index Expanded (SCI-Expanded) 41 Taxonomy of data bases 55 Publishers 62 Taxonomy of publications (2) 65 How to find existing knowledge 67 Outlining a literature review 80 Conclusions 83 Abbreviations 87 References Aarne Mämmelä 12 6

7 Introduction (1) Idea Literature review Problem and hypotheses Experiments/ analysis System (prototype) Theory/paper (new knowledge) Aarne Mämmelä 13 Introduction (2) Literature review is a written well-organized summary of the state of the art (= current technology) with a clearly defined focus should be the second chapter in your doctoral thesis (brief review needed in every original paper in the introduction) a review is written for experts in the field a tutorial is a review written for students (you are not writing a tutorial for your thesis) papers simply collected in your files are not a review! Aarne Mämmelä 14 7

8 Introduction (3) Motivation for literature searches [Sternberg81], [Michaelson90]: full grasp of subject (large picture) show originality of own work list of references for your own paper and thesis ideas for new research Every day more than 1000 papers are published in electrical engineering and electronics (INSPEC), thus you must carefully select what you read the amount of papers is doubled every years Aarne Mämmelä 15 Learning Process [Felder88] Methods of reasoning: Induction (bottom up): from particular to general (the truth not necessarily preserved) Deduction (top down): from general to particular (truth is preserved) Aarne Mämmelä 16 8

9 Taxonomy of publications (1) High reliability, low newness Encyclopedias Monographs Textbooks Review reviews papers Tutorial papers Original journal papers Other original papers - conferences - symposia - workshops Reports Low reliability, high newness Aarne Mämmelä 17 Newness of Knowledge usually to obtain the most recent information it is enough that you read the most recent journals and conference proceedings (technical reports belong to the grey literature) Aarne Mämmelä 18 9

10 Example Original Full Paper Aarne Mämmelä 19 Example Letter (IEE, now IET) Aarne Mämmelä 20 10

11 Example Review Paper Aarne Mämmelä 21 Example magazine paper (tutorial paper) Aarne Mämmelä 22 11

12 Examples of Data Bases Example Internet bookstore Amazon, Exampledatabases IEEE/IET Electronic Library, IEL (user interface is called IEEE Xplore, ieeexplore.ieee.org) Information Services in Physics, Electrotechnology, Computers, and Control (INSPEC, SCOPUS ( Science Citation Index Expanded (SCI-Expanded, see Web of Science, isi01.isiknowledge.com/portal.cgi) Aarne Mämmelä Aarne Mämmelä 24 12

13 Aarne Mämmelä Aarne Mämmelä 26 13

14 Aarne Mämmelä Aarne Mämmelä 28 14

15 Aarne Mämmelä Aarne Mämmelä 30 15

16 Aarne Mämmelä Aarne Mämmelä 32 16

17 Aarne Mämmelä 33 Data bases (2) Aarne Mämmelä 34 17

18 Aarne Mämmelä Aarne Mämmelä 36 18

19 Aarne Mämmelä Aarne Mämmelä 38 19

20 Aarne Mämmelä Aarne Mämmelä 40 20

21 Use of Science Citation Index Expanded (SCI-Expanded) and SCOPUS Aarne Mämmelä Aarne Mämmelä 42 21

22 Aarne Mämmelä Aarne Mämmelä 44 22

23 Aarne Mämmelä Aarne Mämmelä 46 23

24 Aarne Mämmelä Aarne Mämmelä 48 24

25 Aarne Mämmelä Aarne Mämmelä 50 25

26 Aarne Mämmelä Aarne Mämmelä 52 26

27 Aarne Mämmelä Aarne Mämmelä 54 27

28 Largest Databases in Engineering Engineering Village (1969-, partially 1899-) most comprehensive databases in engineering, no citations COMPENDEX (9.2 million), INSPEC (8.8 million), NTIS (2.3 million) altogether 20.3 million abstracts (2005) Web of Science (1986-) Science Citation Index Expanded (SCI-Expanded), Social Sciences Citation Index (SSCI), Arts & Humanities Citation Index (A&HCI) most comprehensive citation search altogether 21.5 million abstracts (2005) SCOPUS (1966-, references 1996-) Chemistry, Physics, Mathematics, and Engineering; Life and Health Sciences; Social Sciences; Psychology and Economics; Biological, Agricultural and Environmental Sciences; General Sciences Competitor of Web of Science, largest database available altogether 28 million abstracts (2006) Aarne Mämmelä 55 Taxonomy of Data Bases (1) 1. Bibliographies best way to find knowledge if available See Science Citation Index, General Search, Document type: Bibliography or Review See SCOPUS, Document type: Review 2. Electronic libraries (whole papers included as.pdf files) IEEE/IET Electronic Library (IEL) from 1950 (user interface is IEEE Xplore), 1.3 million publications, new publications/year ACM Digital Library, portal.acm.org/dl.cfm SPIE Digital Library, spiedl.org Aarne Mämmelä 56 28

29 Taxonomy of Data Bases (2) 3. Abstracts data bases INSPEC from 1969, 8.8 million abstracts, new abstracts/year COMPENDEX from 1969, 9.1 million abstracts, new abstracts/year, much overlap with INSPEC 4. Citation data bases Science Citation Index Expanded (SCI-Expanded) from 1986, 17.5 million abstracts, 1.3 million new abstracts/year, included in ISI Web of Science SCOPUS from 1966, 28 million abstracts See also IEEE Xplore, AbstractPlus, Citing Documents Aarne Mämmelä 57 Taxonomy of Data Bases (3) Other useful data bases include Google Scholar, scholar.google.com (papers) CrossRef within IEEE Xplore (papers) Scientific Literature Digital Library (CiteSeer), citeseer.ist.psu.edu/cs (papers) Google Book Search, books.google.com (books) Project Gutenberg, (books) Barnes & Noble, (bookstore) Abebooks, (used books) Columbia Encyclopedia, (encyclopedia) Acronym Finder, New but in general unreliable information can be found from the Internet, for example Google, (search engine) Wikipedia, en.wikipedia.org (free encyclopedia) Do not in general refer to the Internet in a scientific report or paper (no referee process)! Aarne Mämmelä 58 29

30 Publications in INSPEC and IEEE Xplore Publications/year in INSPEC and IEEE Xplore INSPEC all Publications/year Year INSPEC journal papers INSPEC conference papers IEEE Xplore all IEEE conference papers IEEE journal papers Note. For INSPEC the numbers of the year 2004 are used instead of 2005 (year 2005 not yet complete) Aarne Mämmelä 59 Publications in IEEE Xplore Publications/year in IEEE Xplore Publications/year IEEE Xplore all IEEE conference papers IEEE journal papers Year Aarne Mämmelä 60 30

31 Summary of INSPEC and IEEE Xplore INSPEC data dase (2005) 8.8 million publications new publications/year (2004) doubledin years IEEE Xplore data base (2005) 1.3 million publications new publications/year doubledin 7-10 years In the ratio IEEE conference papers/ieee journal papers was Aarne Mämmelä 61 Publishers (1) 1. Non-profit scientific societies publish best journal and conference papers and some good books, examples Association for Computing Machinery, ACM ( American Physical Society, APS ( Institute of Electrical and Electronics Engineers, IEEE ( International Society for Optical Engineering, SPIE ( Aarne Mämmelä 62 31

32 Publishers (2) 2. Commercial publishers publish good books, examples McGraw-Hill ( Prentice-Hall (vig.prenhall.com) John Wiley & Sons ( note that different publishers are good in different fields, for example in philosophy of science good publishers include Oxford University Press ( Princeton University Press (pup.princeton.edu) Routledge ( University of Chicago Press ( Aarne Mämmelä 63 Publishers (3) Wiley (38) PH (27) MH (23) IEEEP (19) Plenum (8) Kluwer (6) Macmillan (6) AW (5) Dover (5) Nostrand (5) Academic (4) Artech (3) Cambridge (3) Other (26) Derived from the list of references of [Haykin01] (178 books mentioned) Aarne Mämmelä 64 32

33 Taxonomy of publications (2) 1. Scientific books reference books (e.g., encyclopedias), monographs, textbooks, trade books [Day98] 2. Journal and magazine papers review and tutorial papers, original papers 3. Conference, symposium, and workshop papers 4. Technical reports standards (state of the art) doctoral theses patents 5. Newspapers science columns Aarne Mämmelä 65 Journal and magazine papers two general groups original journal papers: full papers, letters review and tutorial papers (often magazine papers) strict writing instructions and peer review process in journal and review papers full paper is a well-rounded treatment of a problem area letters (correspondence) are short papers, published fast, include comments on published papers, corrections, open problems, enhancements of a previous paper magazine papers have less formal writing instructions Aarne Mämmelä 66 33

34 Search according to the publication category (1) bibliographies very useful for finding landmark references theyarefor examplein textbooks (bibliographical or historical notes) review and tutorial papers some IEEE Press Books, in addition to the original papers themselves ( introductions of original papers Science Citation Index, General Search, Document type: Bibliography Aarne Mämmelä 67 Literature Review in an Original Paper Aarne Mämmelä 68 34

35 Search according to the publication category (2) books can be found from library data bases (for example Library of Congress, British Library, Internet bookstores (for example Amazon, Barnes and Noble, Abebooks, book publishers (for example Prentice-Hall, vig.prenhall.org; McGraw-Hill, John Wiley & Sons, book reviews in IEEE Magazines Aarne Mämmelä 69 Search according to the publication category (3) review and tutorial papers for examplein Proceedings of the IEEE (usually the best reviews) and in IEEE Magazines, for example IEEE Communications Magazine (usually tutorials and brief reviews) special issues in other journals using keywords review, overview, survey, or history in databases Science Citation Index, General Search, Document type: Review Aarne Mämmelä 70 35

36 Search according to the publication category (4) original journal and conference papers can be found from databases (IEEE Xplore (IEEE/IEE papers as PDF files, note also the tool Citing Documents in AbstractPlus), INSPEC (abstracts of most papers in electrical engineering), and Science Citation Index) author and subject indices in the last issue of the year newest conference proceedings (best conference papers usually published in journals, but the publication delays are long) for very old papers try IEEE cumulative indices, abstracts journals (INSPEC since 1898, Engineering Index (EI) since 1884), or MathSciNet (since 1940) Aarne Mämmelä 71 Search according to the publication category (5) standards a good source of state-of-the-art system knowledge (although sometimes decisions are political) sold by standardization organizations, for example IEC, IEEE, ISO, ITU, and ETSI you must be an active member in standardization groups to have the newest knowledge (usually through the industry) Aarne Mämmelä 72 36

37 Search according to the publication category (6) patents a form of intellectual property (IP), unlike scientific papers, they limit the use of knowledge, ambiguous writing style much of the information in patents is never published elsewhere patent offices: European Patent Office, United States Patent and Trademark Office, Japanese Patent Office, Free Patents Online, Aarne Mämmelä 73 Search according to the publication category - summary (1) bibliographies are very useful for finding landmark references, they are for example in (1) textbooks, (2) review and tutorial papers, (3) some IEEE Press Books, (4) the introductions of original journal papers, (5) Science Citation Index, General Search, Document type: Bibliography books can be found from (1) library databases ( (2) Internet bookstores (for example or or (3) from Internet book services (for example Google Book Search, books.google.com), and (4) book publishers (for example vig.prenhall.org, review and tutorial papers are often in review journals and special issues, for example in Proceedings of the IEEE (usually the best reviews in electrical engineering) and in IEEE Magazines,seealsoSCI- Expanded, General Search, Document type: Review, or SCOPUS, Document type: Review Aarne Mämmelä 74 37

38 Search according to the publication category - summary (2) original journal and conference papers can be found from databases (1) electronic libraries (society-specific, PDF files available), (2) abstracts databases, for example in Engineering Village (abstracts of most papers), (3) citation indices, for example in Web of Science or SCOPUS, intranet.vtt.fi/siniset/tietpalv/kirjasto/ standards are a good source for state-of-the art system knowledge, but decisions may be somewhat political, sold by standardization organizations, for example IEC, IEEE, ISO, ITU, and ETSI, you must be an active member in standardization groups to have the newest knowledge (usually through the industry) Aarne Mämmelä 75 How to Find Existing Knowledge (1) learn the most important concepts and terms from books and review papers try to understand the historical evolution by using bibliographical notes (read and compare the lists of references, see citation chains) history will improve your understanding of the state of the art historical evolution is always somewhat scattered (overlaps, gaps and conflicts between papers, etc.) read only high-quality landmark papers (learn to know good groups) Aarne Mämmelä 76 38

39 How to Find Existing Knowledge (2) it is better to organize your material according to the relationships of the topics (for example in a taxonomy or classification: separate groups based on some factor common to each, hierarchical structure, special cases) include historical notes as additional information (in the beginning or at the end of your review, usually not in original papers) Aarne Mämmelä 77 Read Papers with a Pen (1) 1. Read first only title, list of authors and abstract 2. Read next introduction, conclusions and references study historical context, what is novel 3. Study organization of the paper organization explained in the end of introduction read titles, definitions of general new terms, see figures read assumptions (system model) and results, numerical examples, conclusions in the text Aarne Mämmelä 78 39

40 Read Papers with a Pen (2) 4. Study details detailed terminology, definitions, proofs implement your own models improve results, write an own paper Aarne Mämmelä 79 Outlining a Literature Review Aarne Mämmelä 80 40

41 Example: General MIMO/OFDM Receiver (1) MIMO, multiple input multiple output OFDM, orthogonal frequency division multiplexing UWB, ultra wideband Time and Frequency Offset Synchronization Narrowband or UWB Channel Beamforming OFDM Demodulator Space Time Coding (inner code) Channel Coding (outer code) Estimation of Antenna Array Coefficients Channel Estimation Source: Markku Kiviranta, VTT Aarne Mämmelä 81 Example: General MIMO/OFDM receiver (2) Selection diversity Beverage et al., 1931 OFDM Doelz et al., 1957 RAKE Price and Green, 1958 MIMO diversity Mellen et al., 1955 Beamforming Applebaum, 1966 BLAST Foschini et al Turbo codes Berrou, 1993 Delay diversity Wittneben, 1991 Space-time codes Tarok et al., 1998 MIMO/OFDM Raleigh and Cioffi, MIMO, multiple input multiple output OFDM, orthogonal frequency division multiplexing BLAST, Bell Labs space time architecture MIMO techniques Space time coding Spatial multiplexing Beamforming Source: Markku Kiviranta, VTT Aarne Mämmelä 82 41

42 Conclusions (1) High reliability, low newness Encyclopedias Monographs Textbooks Review reviews papers Tutorial papers Original journal papers Other original papers - conferences - symposia - workshops Reports Low reliability, high newness Aarne Mämmelä 83 Conclusions (2) Aarne Mämmelä 84 42

43 Conclusions (3) Proceed from general information (encyclopedias, textbooks, reviews) to more specific, learn the terminology first It is best to concentrate on landmark (or seminal) books and papers, which you often see in the lists of references and in bibliographies The most efficient way to find existing knowledge is to use bibliografies (see also reviews in the original papers) Relevant questions are: what is the state of the art and what development lead to it? Write your review by using a classification and the top down approach, take care of the focus Aarne Mämmelä 85 Examples of Good Review Papers S. U. H. Qureshi, Adaptive equalization, Proceedings of the IEEE, vol. 73, pp , September E. Biglieri, J. Proakis, and S. Shamai, Fading channels: informationtheoretic and communications aspects, IEEE Transactions on Information Theory, vol. 44, pp , Oct S. Haykin, Cognitive radio: brain-empowered wireless communications, IEEE Journal on Selected Areas in Communications, vol. 23, pp , February Aarne Mämmelä 86 43

44 Abbreviations ACM, Association for Computing Machinery, COMPENDEX, Computerized Engineering Index, see EI and EI, Engineering Index, ETSI, European Telecommunication Standards Institute, IEC, International Electrotechnical Commission, IEE, Institution of Electrical Engineers (UK), see IET IEEE, Institute of Electrical and Electronics Engineers (USA), IEL, IEEE/IET Electronic Library, ieeexplore.ieee.org, IET, Institution of Engineering and Technology, includes former IEE and IIE, IIE, Institute of Industrial Engineering, see IET INSPEC, Information Services in Physics, Electrotechnology, Computers and Control, see IET and ISI, Institute for Scientific Information, now Thomson Scientific, ISO, International Stardards Organization, ITU, International Telecommunication Union, PDF, portable document format SCI, Science Citation Index, see ISI and isi01.isiknowledge.com/portal.cg SPIE, International Society for Optical Engineering, Aarne Mämmelä 87 References R. A. Day, How to Write and Publish a Scientific Paper, 5th ed. Oryx Press, R. M. Felder and L. K. Silverman, Learning and teaching styles in engineering education, Engineering Education, pp , April S. Haykin, Communication Systems, 4 th ed. Wiley, H. B. Michaelson, How to Write and Publish Engineering Papers and Reports. 3rd ed. Oryx Press, Phoenix, AZ, D. Sternberg, How to Complete and Survive a Doctoral Thesis. St. Martin s Press, Aarne Mämmelä 88 44