GHC13 Banner

Wednesday, February 15, 2012

Learning to Research, Part 4 Learning to Search, Narrow , and Visualize

A little reminder before we start. You need to have an off campus library log in to be able to search databases and catalogs for things restricted to UCF student use such as thesis and dissertation text, ebooks, and journal articles.

Part 3 Continued
                                                               i.      Research Methods and Data Sources - links you to a page with useful information for graduate students.
1. Read it

Searching the existing Dissertation Database

  1. Decide what your generic area is, in this case DSP. You need somewhere to start looking. Before you get more specific you want to know all possible people and databases that you could look for to gain information on what is worth while.
    1. The first place you want to check for topics is the library dissertation and thesis titles. The reason is that previous UCF students will be a lot more willing to give you a place to start. Especially if they just graduated, they may even give you code from their dissertation.
  1. Attempt to use review journals or databases to let you know what current and up and coming topics are, for example:
  1. Is there something more specific such as an intersecting field, this can help narrow your topic.

Example Thesis and Dissertation Search
I want to know what people are doing in the area of speaker recognition across the world. So I go to the ProQuest database search and look in the thesis and dissertation database searching on speaker recognition, and I limit my search to English documents.

49,703 results are returned! Here is a histogram of the papers released by decades, looks like the first paper returned was in 1912 and we even have some prerelease listings for next year, 2013. We want to know what the latest dissertations are right? So first let us narrow our search to the last two decades. How many papers is that?

45,316 results still remaining! Not small enough, lets take a look at that updated histogram,

and while we are at it lets look at the list of subject areas we are searching. The list is verrrry long, I've narrowed mine to Computer Science and Electrical Engineering. Let's see whats remaining. 1563 results! That is a lot more manageable.

From this subset I've taken the list of author keywords and created a wordle visualization of these keywords, take a look

Do you see just under Machine Learning there is the keyword signal processing? Lets narrow our search to the keywords signal processing as well.

And now we have 18! Yes! Lets look at these 18 titles and authors. Remember that these are dissertations or thesis so the authors have graduated from school in the last 10 years and are probably now the leading researcher in these areas. Also these dissertations have required multiple publications before they were written.

Here are our authors, titles, schools and years:

Title Authors School Year Number
Robust and efficient techniques for speech recognition in noise  Sarikaya, Ruhi; Hansen, John H L  Duke University, 2001 2001 3031012
High range resolution radar target classification: A rough set approach  Nelson, Dale; Starzyk, Janusz A  Ohio University, 2001 2001 3019554
Enhancement and recognition of whispered speech  Morris, Robert; Clements, Mark A  Georgia Institute of Technology, 2003 2003 3110442
Signal processing strategies for better melody recognition and improved speech understanding in noise for cochlear implants  Kasturi, Kalyan; Loizou, Philipos C  The University of Texas at Dallas, 2006 2006 3238589
Multi-classifiers and decision fusion for robust statistical pattern recognition with applications to hyperspectral classification  Prasad, Saurabh; Bruce, Lori M; Fowler, James E; Younan, Nicolas H; Du, Jenny Q  Mississippi State University, 2008 2008 3331332
Model-based speech separation and enhancement with single-microphone input  Lee, Siu; Ching, Chung  The Chinese University of Hong Kong (Hong Kong), 2008 2008 3348874
Speech processing and modeling using a non-linear time-frequency algorithm  McNamara, David  Clarkson University, 2008 2008 3340051
A sequential algorithm for biological event detection using statistical nonstationarity  DiCecco, John  University of Rhode Island, 2008 2008 3328721
A model of head-related transfer functions based on a state-space analysis  Adams, Norman; Wakefield, Gregory H  University of Michigan, 2008 2008 3304909
Signal processing meets computer vision: Overcoming challenges in wireless camera networks  Yeo, Chuohao; Ramchandran, Kannan  University of California, Berkeley, 2009 2009 3383584
Distributed multichannel processing for signal enhancement  Trawicki, Marek  Marquette University, 2009 2009 3357974
Spectral refinements to speech enhancement  Charoenruengkit, Werayuth; Erdol, Nurgun  Florida Atlantic University, 2009 2009 3351583
Applications of sensor arrays in acoustic and seismic signal processing  Xie, Peng; Grant, Steven L  Missouri University of Science and Technology, 2009 2009 3365041
Nonstationary time series modeling with applications to speech signal processing  Rudoy, Daniel; Wolfe, Patrick J  Harvard University, 2010 2010 3435443
Dimensionality reduction and fusion strategies for the design of parametric signal classifiers  Kota, Srinivas; Gupta, Lalit; Bhattacharya, Bhaskar; Botros, Nazeih; Choudhary, Ruplal; et al  Southern Illinois University at Carbondale, 2010 2010 3440304
Acoustic MEMS array embedded in a scalable real-time data acquisition and signal processing platform  Turqueti, Marcos; Saniie, Jafar  Illinois Institute of Technology, 2010 2010 3435825
Use of acoustic analysis to develop, test and optimize cochlear implant sound processing  Won, Jong; Rubinstein, Jay T  University of Washington, 2010 2010 3431627
Architecture and analysis for next generation mobile signal processing  Woh, Mark; Mudge, Trevor N  University of Michigan, 2011 2011 3477118

Remember this is not a perfect science, finding the right things is an art, so do these look like what we were looking for? Yes and no depending on your specific area, but its close.

More on searching next time but for now let me point out a good data visualization tool, just be sure to read the rules before using it

Here is a visualization of the data we found:
take a look at the interactive version on many eyes if you like

Until next time!