From: Tim Paek
Sent: Friday, February 12, 2010 12:45 AM
To: 'principal@kcl.ac.uk'; 'jan.palmowski@kcl.ac.uk'; 'keith.hoggart@kcl.ac.uk'; 'chris.mottershead@kcl.ac.uk'
Subject: Letter in support of Jonathan Ginzburg
Dear Professor Rick Trainor, Professor Jan Palmowski, Professor Keith Hoggart and Mr. C. Mottershead,
I am the current President of the Special Interest Group on Discourse and Dialogue (SIGDIAL), a special interest group of the Association for Computational Linguistics (ACL) and the International Speech Communication Association (ISCA). In case you are not familiar with our organization, each year we host the premiere venue for research in discourse and dialogue, encompassing everything from reasoning about beliefs, intentions and actions in dialogue interaction to the practical details of building an intelligent system capable of engaging users in conversation. In short, our venue captures everything that falls under the scope of “Applied Logic and Theory of Computation” (ALTC) and then some. We pride ourselves in bringing together researchers focused on theoretical modeling with those developing real-time conversational agents. Our venue is held in cooperation with the Association for the Advancement of Artificial Intelligence (AAAI) and our research papers attract the attention of not only academia but also industry (e.g., AT&T) and trade organizations (e.g., AVIOS which co-runs SpeechTek).
I am writing this letter in support of Dr. Jonathan Ginzburg. As a valued member of the SIGDIAL community, Dr. Ginzburg’s work has been influential in countless ways. I will outline a few from the perspective of an industry lab researcher who builds dialogue technologies for Microsoft (I helped to ship Voice Command 1.6, Vista Speech Recognition, Bing Mobile, and the upcoming Windows 7 phone). Dr. Ginzburg’s work on Type Theory with Records established a basis for the Information State Update approach to dialogue management, which has been used widely in spoken dialogue systems (i.e., conversational agents or intelligent systems) across the globe. Dr. Ginzburg’s work on clarification and contextual updates has challenged and inspired researchers, such as myself, who work on error-handling and misunderstandings. Although machine learning approaches have gained in popularity over the years, most researchers taking this approach (myself included) recognize that stochastic optimization and control theory can only take us so far. We need to integrate the kinds of logics Dr. Ginzburg has been innovating for decades to get closer to the dream of artificial intelligence. Dr. Ginzburg’s influence extends beyond research. As you are probably aware, he has recently played a vital role in spearheading an online journal, Dialogue and Discourse, of which he is currently Editor-in-chief. In many ways, Dialogue and Discourse is THE journal for SIGDIAL, and we look forward to seeing it flourish under Dr. Ginzburg’s leadership.
In both his research and professional activities, Dr. Ginzburg epitomizes ALTC. To contend that his research somehow fails to fit the ALTC is truly a travesty. I hope and trust that any committee formed to evaluate his fit into ALTC would do its due diligence in seeking out the opinion of researchers whose primary expertise is in applied logic with respect to conversational interaction. If you need any such references, I would be delighted to provide them. Please note that Dr. Ginzburg did not in any way influence me to write this letter. I wrote this because I have always admired Dr. Ginzburg’s research, ever since I was a graduate student. I still look to his research for vision and inspiration.
If you have any questions about this letter, please do not hesitate to contact me.
Sincerely,
Tim Paek
Researcher
Machine Learning and Applied Statistics
Microsoft Research
http://research.microsoft.com/en-us/um/people/timpaek/