Bo Pedersen

See http://www.linkedin.com/in/bopedersendc for recommendations, etc.

Education

Cornell University (2003-2009), Department of Psychology, MS Cognitive Psychology.

Aarhus University, Denmark (1988-1996), MS in Computer Science. Thesis: Parsing Natural Language with Harmony Functions and Canonical Syntax.

Aarhus University, Denmark (1989-1992), Subsidiary subject (Sidefag) in Information Science (1.5 years full time study). Special arrangement for Computer Scientists with course work equivalent to a full BA in Information Science.

Aarhus University, Denmark (1989-1992), Minor in mathematics (1 year full time. Same initial curriculum as for mathematics majors).

Experience

Data Scientist, Rosetta Stone, Harrisonburg, VA, 2010-2013.

As a Data Scientist at Rosetta Stone I've been involved in the analysis of the vast amounts of data that we collect through the user interface, and I have used matrix decomposition, clustering, visualizations, etc. at scale to answer questions like: Are the users learning anything? Are they frustrated? Is this new component/initiative working?. More than often I have been the first person in the organization to do so, and the one to get the data out of the databases, clean it up and normalize it, and more than often I've been faced with data that were incorrect, insufficient or just with very little signal, and it's been part of my job to push to correct or change what we were collecting. And yes, of course I have been using hadoop, hdfs, noSQL as well as SQL, shell scripts, c++ and my old friend Matlab..

However, the most challenging part of my job has been to analyze data for products that we were planning to build and obviously had no user data for yet, and then try to use internal and external sources to create a franken-data-set, in order to test predictability, user experience, and the overall sanity of the planned product.

Highlights:

Devised a generalized matrix decomposition framework to predict user behavior, and suggest appropriate content for users, and implemented the highly scalable c++ version that went into production.
Created a framework that could make tight predictions of users standardized scores using non-expert judgments on Amazon Mechanical Turk. Patent currently being considered by the legal department.
Created a stochastic model that predicted how fast the user population would consume new content.
Created a javascript vector graphic visualization tool for large data sets, where the end-user could manipulate and re-filter the data without making new server-side requests.
Devised a hybrid matrix model that could make recommendations in a variety of different ways from 3 matricies: User/Content, Content/Words, User/Words.
Created a prototype of a new mobile app that has been scoped for production.

Software Engineer, Project Manager, Centre for Language Technology, University of Copenhagen, 2000-2003.

Software engineer on the OntoQuery project (funded by the Danish Research Agency), where we developed a prototype for content-based information retrieval of domain knowledge. My function was to handle low level processing such as tokenization, tagging, etc. http://ontoquery.dk
Project manager on "defisto", a "smart" search engine for a number of text and dictionary resources in Danish. Online, serving the Danish internet community since 2003 http://cst.dk/cgi-bin/sto/defisto
Responsible for setting up, administering, handling data exchange in Oracle for the dictionary project STO (funded by the Danish Ministry for Information Technology and Research). In this function I also programmed a web spider to collect domain specific words for the dictionary, and I contributed to the Danish version of the open source project Aspell. http://cst.dk/sto/uk
Software Engineer on the Staging project (funded by The Danish Research Council), a multimodal interface where I was responsible for programming a little "virtual environment" - a small Dynamical System that handles language events like "to give" without an explicit "plan" but with small changes in the dynamics, 2001. http://babbletower.dk/staging.mp4
Project Manager on the ToolBox project, collecting information about existing software at Centre for Language Technology and giving advice on further development, 2001. http://www.cst.dk/online/uk/index.html
Responsible for system security. Sat up a firewall on Linux and filters for virus and spam (RedHat, IpChains, Batemail, SpamAssassin). Also responsible for mail and web server, 2001-2002.
Participated in the Nomen Nescio - named entity recognition (funded by Nordic Academy for Advanced Study, NorFA), where I modified the Brill Tagger to be used for named entity recognition, 2001.

Software Engineer, Project Manager, EDB Gruppen A/S (IBM owned Danish software house), 1999-2000.

Commercial Project Manager for tailoring a DOPS (hour registration) system for a costumer.
Collection and development of mobile internet solutions within the organization.
GPS consultant. I helped Vejdirektoratet (Danish Road Directorate) program a system that could communicate and handle data from 2 GPS receivers for the positioning of offshore oil platforms.

Software Engineer, Project Manager, Oodac ApS, 1997-1999.

Developed a parser of time expression for the Event database at Danish newspaper Jyllandsposten, so that given a complex descriptions of an event (concerts, museums), an exact time entry can be made in the database and be ready for newspaper production or online query, and responsible for the xml module that produces the xml for both media. Developed web interface for the administration of banner ad's. http://babbletower.dk/timeparse
Project Manager on a car developed for The Danish Road Directorate for data collection. The solution controlled and collected data from 3 moving cameras, GPS, a gyroscope with the help of a touch screen and speech recognition. http://www.smalltouch.com/pdf/smallPOS_brochure.pdf
Project Manager on the development of a new touch screen to be used for voting in The City Hall of Copenhagen. Coordinated design, hardware, programming (of which I did a substantial amount myself). http://smalltouch.com/touchm2.html

Teaching Experience

Teaching Assistant at Cornell University (Office hours, review sessions, grading, occasional lecture)

PSYCH 205, Perception (James Cutting), Fall 2008.
PSYCH 350, Statistics and Research Design (Jennifer Schwade), Fall 2007.
PSYCH 101, Introduction to Psychology (Ann Lynn), Summer 2007.
PSYCH 325, Adult Psychopathology (Harry Segal), Spring 2007.
PSYCH 102, Introduction to Cognitive Science (Michael Spivey), Fall 2006.
PSYCH 102, Introduction to Cognitive Science (Florencia Reali), Summer 2006.
PSYCH 416, Modelling Perception & Cognition (Michael Spivey), Spring 2006.
PSYCH 102, Introduction to Cognitive Science (Breton Bienvenue), Fall 2005.
PSYCH 102, Introduction to Cognitive Science (Rick Dale), Summer 2005.
PSYCH 214, Cognitive Psychology (Shimon Edelman), Fall 2004.

Instructor, International Computer Driving Licence (ICDL), entry-level and advanced modules (Microsoft Office). I also taught short courses in basic computer use for immigrants, and created my own 6-week intensive course "CyberTramp", a pragmatic guide on how to create an identity on the internet with special focus on free resources, Silkeborg Daghøjskole, 1996-1997.

Middle-school teacher, Introduction to Computers, Havvejens Ungdomsskole, Egå, 1996. Teaching the kids how to make a version of the pacman game in Visual Basic.

Ad hoc instructor for language translation software, Department of Psychology, Aarhus University, 1995.

Publications, Presentations, etc.

Bo Pedersen, Florencia Reali, Michael Spivey, Kelly Conway The Role of Set Size in Linguistically Mediated Visual Search (in preparation)
Visual search source code:

Nadya Dich, Bo Pedersen Native Language Affects Spelling in English as a Foreign Language: a Time-Course Analysis of Proficient EFL Spelling, Canadian Journal of Applied Linguistics (in press)

Bo Pedersen A Dynamical Systems Reformulation of the Normalized Recurrence Algorithm poster at the 33th Annual Meeting of the Cognitive Science Society, CogSci2011 Proceedings, 2011.

Bo Pedersen Crossmodal Entropy Transfer, poster at the 29th Annual Meeting of the Cognitive Science Society, CogSci2007 Proceedings, 2007.
Crossmodal Transfer Entropy source code:

Bo Pedersen Interleaved and Alternating Learning of Dependent and Independent Stimuli, poster at the North East regional Sigma Xi conference, 2007.

Bo Pedersen, Michael Spivey Offline tracking of eyes and more with a simple webcam in Proceedings of the workshop "What have eye movements told us so far, and what is next?", 28th Annual Meeting of the Cognitive Science Society, 2006.
Argus source code:

Bo Pedersen Visual Cues in Connectionist Models of Language Acquisition, poster at the 28th Annual Meeting of the Cognitive Science Society, CogSci2006 Proceedings, 2006.

Bo Pedersen, Shimon Edelman, Zach Solan, David Horn, Eytan Ruppin Some Tests of an Unsupervised Model of Language Acquisition, in Proceedings of COLING-2004 Workshop on Psycho-computational Models of Human Language Acquisition, Geneva, Switzerland, 2004.

Bo Pedersen, Shimon Edelman, Zach Solan, David Horn, Eytan Ruppin Comparing Acceptability in Magnitude Estimation Tests of an Unsupervised Model of Language Acquisition, poster at the 26th Annual Meeting of the Cognitive Science Society, CogSci2004 Proceedings, 2004.

Bo Pedersen, Shimon Edelman, Zach Solan, David Horn, Eytan Ruppin Acceptability Experiments in ADIOS, poster at the Cornell Symposium on Language Universals, May 1-2, 2004.

Bo Pedersen The representation of motion in language, presentation at the Empirical Methods in Cognitive Linguistics Workshop, Cornell University, May 2-4, 2003.

Nicolai H. Sørensen, Bo Pedersen: Demonstration of the sto database (A large computational lexicon for Danish) and defisto interface at the Tenth EURALEX International Congress, Copenhagen, August 14-17, 2002.

Patrizia Paggio, Bart Jongejan, Bo Pedersen: Demonstration of the multimodal Staging environment at the MUMIN (A Nordic Network for MUltiModal INterfaces) Workshop, Tampere, Finland, November 15-23, 2002.

Bo Pedersen Automatisk navnegenkendelse ved hjælp af ordklasser (Named Entity Recognition with Word Classes) in Holmboe, Henrik (eds) Nordisk sprogteknologi - Nordic Language Technology, Museum Tusculanum Press, University of Copenhagen, 2001.

Bo Pedersen The Depth of Different Connectionist Systems and the Consequences for Semiotics Proceedings of 1st International Workshop on Computational Semiotics. Leonardo da Vince University, Paris, 1997.

Bo Pedersen Sentence Imagery: The Transport of Full Sentence Meaning in Open Cognitive Schemas in Colloque CISL 97, L'image dans le langage et dans les non-langages, Papers From the Colloquium, Published by the Center for Semiotic Research, University of Aarhus, 1997.

Bo Pedersen Maskinoversættelse ved hælp af billedskemaer, kanonisk syntaks og dynamiske systemer (Machine Translation with Image Schemas, Canonical Syntax and Dynamic Systems) In Jensen, Per Anker, Jørgensen, Stig. W., Hørning, Annette (eds), Danske ph.d.-projekter i datalingvistik, formel lingvistik og sprogteknologi, Kolding, 1997.

Bo Pedersen Joining Icon, Symbol, and Index in one Connectionist Substrate, in Peter B. Andersen, Mihai Nadin and Frieder Nake (eds) Informatics and Semiotics, Dagstuhl Seminar Report 135, Schloss Dagstuhl, Internationales Begegnungs- und Forschungszentrum fur Informatik, 1996.

Shimon Edelman, Bo Pedersen, review of Linguistic Evolution through Language Acquisition (Ted Briscoe, ed., Cambridge University Press, 2002), Journal of Linguistics vol. 40(2):14-18, 2004.

Bo Pedersen Cognitive, More Cognitive, Most Cognitive, review in Brandt, P.A. (eds) Dynamiske modeller (Dynamic Models), Almen Semiotik 14 (Journal of General Semiotics), Aarhus University Press, 1998.

Bo Pedersen The horse raced past the barn fell, review in Brandt, P.A. (eds) Væren og betydning (Being and Meaning), Almen Semiotik 13 (Journal of General Semiotics), Aarhus University Press, 1997.

Bo Pedersen The stemma editor and some theoretical considerations in Evaluation of the Center for Semiotic Research, Danish National Research Foundation, 1997.

Bo Pedersen Informatics and Semiotics, review in Brandt, P.A. (eds) Etik og passion (Ethics and Passion), Almen Semiotik 11-12 (Journal of General Semiotics), Aarhus University Press, 1996.

Bo Pedersen The First International Summer Institute in Cognitive Science. Buffalo 1994., review in Brandt, P.A. (eds) Filosofi og semiotik (Philosophy and Semiotics), Almen Semiotik 9-10 (Journal of General Semiotics), Aarhus University Press, 1995.

Ad hoc reviews for: Cognitive Science Journal, Image and Vision Computing Journal, The Annual Conference of the Cognitive Science Society, The International Conference on Cognitive Science, The European Cognitive Science Conference.

Translations:

Substantiver og Verber Translation of Ronald W. Langacker's Nouns and Verbs. In P. F. Bundgård, J. Egholm, M. Skov (eds) Kognitiv Semiotik: En Antologi om Sprog, Betydning og Erkendelse, 253-293, Copenhagen, Haase og Søn, 2002.

Abstraktionens Former ifølge Kierkegaard Translation of Darío González' Modos de la abstracion segun Kierkegaard (Modes of abstraction according to Kierkegaard). Translated for Almen Semiotik, but never published, 1992.

Member/Fellow/Scholar -ships

Thanks to Scandinavia Scholarship, 2003-2004.

President of the Scandinavian Club at Cornell, 2004-2006.

Grant from Denmark's Electronic Research Library for the DEFiSTO web interface for the sto (large computational lexicon for Danish) and corpora search engine at the Centre for Language Technology, University of Copenhagen, which Nicolai H. Sørensen and I applied for (officially with Anna Braasch as the principal investigator but with me as the project manager), 2002.

Maximum points and a special mention in the "Best on the net" evaluation for high usability for the website cst.dk, 2001-2002.

Member of the Cognitive Science Society since 2004.

Registered Linux user #7844 (counter.li.org), since 1994.

Member of the Basic Research Group in General and Dynamic Semiotics at the Center for Semiotic Research, Aarhus University, 1993-1998. Funded by the Danish National Research Foundation.

Member of the board in Smalltouch ApS, 1999-2002.

Chair and member of the board in Ejerforeningen Vesterport 8, Aarhus, Denmark, 1996-2001.

Cornell Cognitive Studies Travel Grant for: The International Conference on Computational Linguistics, Geneva, 2004. 28th Annual Meeting of the Cognitive Science Society, Vancouver, 2006. Cornell University Graduate School Travel Grants for: The International Conference on Computational Linguistics, Geneva, 2004. 26th Annual Meeting of the Cognitive Science Society, Chicago, 2004. 28th Annual Meeting of the Cognitive Science Society, Vancouver, 2006. Computer equipment from Michael Spivey's NIH grant R01 MH63961 "Mutual Influences Across Language and Vision" and travel support for Lake Ontario Visionary Establishment, Perception/Cognition conference, Niagara Falls, Ontario, 2005, 2006, 2007. Summer support from Shimon Edelmans's U.S.-Israel Binational Science Foundation grant "The Emergence of Syntactic Structures and Their Representation", 2004. Travel grant from the National Science Foundation for the Empirical Methods in Cognitive Linguistics Workshop, Cornell University, May 2-4, 2003. Travel grants from the Nordic Academy for Advanced Study (NorFA) for the Nomen Nescio seminar, Fefor, Norway, 2003; The first MUMIN network meeting and Ph.D. course in Tampere/Helsinki, Finland, 2003. Scholarship from Centre Alexandre KOYRE for the colloquium "Sciences and Languages in Europe", Ecole des Hautes Etudes en Sciences Sociales, Paris, 1994. Computer equipment grant from Center for Semiotic Research, Aarhus University, 1994. Travel grants from the Computer Science department, Aarhus University, for the The First International Summer Institute in Cognitive Science, Buffalo 1994. Travel grants from the Center for Semiotic Research, Aarhus University, funded by the Danish National Research Foundation, for: The First International Summer Institute in Cognitive Science, Buffalo 1994. The First International Cognitive Technology Conference, Hong Kong, 1995. COLING 96, The 16th International Conference on Computational Linguistics, Copenhagen, 1996. Dagstuhl Seminar 9608 on Informatics and Semiotics, at The International Conference and Research Center for Computer Science, at Schloss Dagstuhl, 1996. ESSLLI'97, 9th European Summer School in Logic, Language and Information, Aix en Provence, France, 1997. Stage di semiotica at Centro Internazionale di Semiotica e Linguistica Università degli Studi di Urbino, Italy, 1992-1998. Danske ph.d.-projekter i datalingvistik, formel lingvistik og sprogteknologi, Kolding, 1997. The Roman Jakobson Centennial Symposium, Copenhagen, 1998. The Structure of Signs, Things, and Language Symposium, Rolighed, Skodsborg, 1995. The Structure of Intentionality Symposium, Rolighed, Skodsborg, 1996. Dynamisk Semiotik konference (Dynamic Semiotics conference), Sostrup Slot, 1992. Travel grant from the Copenhagen Business School for participation in the Ph.D. course Oversættelsesteori (Translation Theory), 1996.

Other

Developed a couple of toolboxes in Matlab (Eye-tracking, Visual Search) that are used in a handful of labs across the US. http://babbletower.dk/shared/

Programmed 2 projects for the Johnson School of Business at Cornell for collecting data online. (Matlab, bash, OS X).

Designed and administered a number of websites throughout the years (including one award winning).

Native Danish speaker, fluent in English, read/write German, and speak it too if you press me hard.

European Summer School in Logic, Language and Information, Aix en Provence, France, 1997.

First International Summer Institute in Cognitive Science, Buffalo, NY, 1994.

Oracle 8i Enterprise DBA Part 1A: Architecture and Administration, 2001.

Java, 2 day Course, 1998

Emergent grammar parser for Danish with Robert Freeman, 2002.

Marked up a Danish poetry collection in a construction grammar manner and sat up a website that create novel poems every time the site is accessed: babbletower.dk. Personal project, 2001. Adopted by The Danish Online Dictionary in 2004 for their online poetry workshop.

Developed the software for testing the perception of emotions in facial expressions at the Center for Semiotic Research, Aarhus University, 1995.

Developed a graphical editor for the semiolinguistic (stemma/stemmatics) theory at the Center for Semiotic Research, Aarhus University. In cooperation with Jens Henrik Skovgaard and Per Aage Brandt, 1994 (Objective-C). http://babbletower.dk/stemma

Repeated David Rumelhart's neural network experiment with past tense verbs for Danish. In cooperation with Steen Langballe Nielsen and with advice from Brian Mayoh, Chris Sinha, Kim Plunkett, David Rumelhart, 1994. Unpublished paper: Den nervøse datid - Om modelleringen af børns indlæring af danske datidsverber ved hjælp af kunstige neurale net (The Nervous Past Tense - On The Modelling of Childrens Acquisition of Danish Past Tense Verbs with Artificial Neural Networks).

Project on symbolic differentiation in Lisp and Comal 80, 1987

Erhvervspraktikant at LEGO Group where I programmed a little employee registration system, 1983

References

Ilse Ackerman, Data Scientist (Manager), Rosetta Stone, ilse.ackerman@gmail.com

Michael J. Spivey, Professor of Cognitive Science, University of California, Merced, (209) 228-7742, spivey@ucmerced.edu http://faculty1.ucmerced.edu/spivey/

Per Aage Brandt, Professor, Case Western, (216) 368-4753, peraage.brandt@case.edu http://www.case.edu/artsci/cogs/brandt.html

Patrizia Paggio, Senior Researcher, Centre for Language Technology, +45 3532 9072, paggio@hum.ku.dk http://cst.dk/patrizia/paggio/uk/index.html

Jens Holtsmark, CEO, owner, Holtsmark A/S. +45 2025 4060, jens@holtsmark.com http://www.holtsmark.com