Beruflich Dokumente
Kultur Dokumente
• Machine Learning
http://www.kluweronline.com/issn/0885-6125
• Music Perception
http://www.ucpress.edu/journals/mp/
• Psychology of Music
http://pom.sagepub.com/
• IEEE Transactions on Pattern Analysis & Machine
Intelligence
http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber
=34
• IEEE Transactions on Multimedia
http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber
=6046
Getting the information:
Conferences
• Int. Computer Music Conference (ICMC)
• Int. Conf. On Digital Audio Effects (DaFX)
• Int. Conference on Music Information Retrieval (ISMIR)
• IEEE Int. Conf. On Acoustics, Speech and Signal
Processing (ICASSP)
• Computer-Human Interaction (CHI)
• New Interfaces for Musical Expression Conference
(NIME)
• ACM-Multimedia
• Sound and Music Computing
Be careful with papers from obscure or
very local conferences:
quality standards can be too low
Getting the information:
Patent databases
Pablo García will tell you why/what/how soon!
Otherwise, consider the option of having hit a goldmine (new and original
topic)
– BUT if you perseverate on the topic you will not be given the option of
“walking on the shoulders of giants”
– Ask other senior researchers (your supervisor could have turned crazy)
Getting the information:
Keeping up-to-date
• [Music-DSP]
http://www.music.columbia.edu/cmc/music-dsp/
• [Auditory]
http://www.auditory.org/
• [Music-IR]
http://www.ismir.net/
• [Weka]
https://list.scms.waikato.ac.nz/mailman/listinfo/wekalist
• [Semantic Web]
http://lists.w3.org/Archives/Public/semantic-web/
• [SMC]
http://smcnetwork.org/
Organizing the information
• By source
– It allows you to develop how one researcher or group
of researchers has contributed to the field (paperwise
structure: Smith (1999), then Smith et al. (2000), then
Smith (2001) –one paragraph per researcher-)
• By topic
– It allows you to cover all of the contributions, by
different researchers to one topic, problem or key
area of knowledge (conceptwise structure: you build a
map of the subtopics included in your topic and flesh
the map mixing dates and authors)
• By method/technique/algorithm/approach
• Chronologically
Organizing: grouping your sources (I)
Organizing: grouping your sources (II)
Organizing: Summarizing
• Try to determine which variables or
dimensions make papers similar or
different
• Build tables summarizing information
• Draw flowcharts
Organizing: asessing the rigour of
your sources
Organizing: asessing the rigour
of your sources
• Asessing the impact of authors, journals,
articles:
– You can get or track authors’s impact indexes
and more relevant info using
• Publish or perish (http://www.harzing.com/pop.htm)
• Scimago (http://www.scimagojr.com/index.php)
• Arnetminer (http://www.scimagojr.com/index.php)
• Google Scholar (http://scholar.google.com/)
• Science Citation Index (several entry points –
Library, for example)
Organizing: Evaluating the work done
You need to assess the work done in order to establish:
1. What are the significant points of agreement between
articles?
2. Where the research disagrees, is one researcher more
conclusive than another?
3. How previous work has left a gap, because of either
inadequate assumptions, inconclusive findings, poor
methodology, unclear presentation of results,
unavailability of recent technologies, etc;
4. How previous research will be applied in a new context;
5. How general disagreement or different methods, results
or approaches create a need for a solution.
6. How can you fit the articles together to build a logical
argument that furthers your purpose.
Organizing the information
• Breadth / Depth tradeoff
• From the review, you should be able to
spot on:
– The “core” papers (to be read and
summarized)
– The “interesting” papers (to be succintly
summarized in one or 2 paragraphs)
– The garbage papers (the do exist!): to be
omitted or, better, criticised
Writing the document
• http://www.upf.edu/bibtic/recursos/treaca/redacteu.html
• Be aware about formatting requirements (by the UPF, by
the potential publisher)
• Which is the style for your MSc Thesis? (see the section:
template)
– http://www.upf.edu/bibtic/en/guiesiajudes/tesis/dina4.html
• Get familiar with IEEE, ACM, and APA citation styles
• Use a bibliography manager:
– Zotero http://www.zotero.org/
– JabRef http://jabref.sourceforge.net/
– Mendeley http://www.mendeley.com/
– See for comparison tables:
http://en.wikipedia.org/wiki/Comparison_of_reference_
management_software
Writing the document
• Start with an outline,
– then flesh it up with a summary for each section
– then start adding the main content to each section
– after some writing your outline should be redone, indeed!
• Write, write, write but write for a reader (not for yourself)
– Don’t waste time looking at an empty page
– Don’t stare at the ceiling
– Don’t wait to feel inspired
– Don’t procrastinate
– Get some help from productivity analysis tools (e.g., RescueTime)
• Don’t be critical with your own writing until a draft is
finished
– Re-read, correct, delete, rephrase (minimum 8000 words should be
used) in order to make it understandable
Writing the document
• Be sensitive to writing style issues (yes,
engineers are not poets, but have to communicate
ideas in the most efficient way -> Style is for that:
exposing, connecting, convincing)
• Give your first draft to your supervisor or office
mate
– Give it at the proper time, not too early
(ununderstandable, unconnected) but not too late (when
you have committed and fallen in love with your view
and text)
– Do English and proofreading on your own devices
(you don’t want your supervisor waste her precious time
on that! –She neither does!)
Writing the document
Use evidence
• Your statements have to refer to several sources when
making your point (this gives them credibility, validity).
Select those sources giving you the most credibility (e.g. a
solid author is preferred over Wikipedia entries, a journal
article is preferred over a MSc Thesis).
Be selective
• Select only the most important points in each source to
highlight in the review. Exhaustivity is not needed in a
MSc thesis but it is in a PhD thesis.
Use quotes sparingly
• Some short quotes here and there are okay, though, if you
want to emphasize a point, or if what the author said just
cannot be rewritten in your own words. Keep quotation for
certain terms that were coined by the author, not common
knowledge, or taken directly from the study.
Writing the document
Summarize and synthesize
• Remember to summarize and synthesize your sources
within each paragraph as well as throughout the review.
Connect the findings to your own goals, use them to
backup your position.
Keep your own voice
• While the literature review presents others' ideas, your
voice (the writer's) should remain front and center. Use
the sources to support what you want to present. If you
fail on that, then you can probably be “building Castles in
Spain”, or selling “vapourware”.
Use caution when paraphrasing
• When paraphrasing a source that is not your own, be
sure to represent the author's information or opinions
accurately and in your own words. Double-check that the
ideas you present as “original” cannot be traced back to
any of your sources.
A simple example
The last
paragraph is not
typical in a s.o.t.a.
unless you are
ending it and
presenting your
work in summary
Add value to your state of the art
• It can provide a needed update (if the topic has
not been summarized in the last 3 years)
• It can provide a tutorial or introductory reading to
other students
• It can glue or link different pieces of knowledge,
disciplines, techniques, etc.
• It can be valuable as a “stand-alone” article for a
journal
• Convince a “big name” in your topic area to
participate in the article (it increases the quality
and the chance to get published)
Some examples
• Journal of New Music Research Volume 32,
Issue 1, 2003 contains several state-of-the-art
reviews (by Emilia, Fabien Gouyon, or Perfe)
• http://www.informaworld.com/smpp/title~d
b=all~content=g714018053
• Please, send me additional useful
pointers, missing info, suggestions, etc. to
improve the ppt for future editions!!!
• Thanks!!