For the upcoming TSD, the following outstanding set of invited speakers with various expertise covering speech modeling, acoustic-phonetic decoding, dialogue systems, and semantics agreed to give their respective pieces of speech:
Hynek Hermansky (The Center For Language and Speech Processing, Johns Hopkins University, Baltimore, USA):
Long, Deep and Wide Artificial Neural Nets for Dealing with Unexpected Noise in Machine Recognition of Speech.
Abstract: Most emphasis in current deep learning artificial neural network based automatic recognition of speech is put on deep net architectures with multiple sequential levels of processing. The current work argues that benefits can be also seen in expanding the nets longer in temporal direction, and wider into multiple parallel processing streams.
Torbjörn Lager (Department of Philosophy, Linguistics and Theory of Science, University of Gothenburg, Sweden):
Statecharts and SCXML for Dialogue Management.
Abstract: The World Wide Web Consortium (W3C) has selected Harel Statecharts, under the name of State Chart XML (SCXML), as the basis for future standards in the area of (multimodal) dialog systems. In this talk, I give a brief introduction to Statecharts and to SCXML, show what it can do (and not do) for someone wanting to use it for dialogue management, describe its relation to other dialogue system components, explain its possible use as a meta-dialogue manager (i.e. as a manager of dialogue managers), as well as some ideas for how to compile other dialogue management languages (such as VoiceXML) into SCXML. If time permits, I will also give some pointers to existing SCXML implementations.
Ralf Steinberger (European Commission – Joint Research Centre, Ispra, Italy):
Multilingual Media Monitoring and Text Analysis – Challenges for Highly Inflected Languages.
Abstract: The European Commission’s Europe Media Monitor (EMM) family of applications helps users monitor multilingual written online media for information on a wide variety of subject domains. Apart from gathering an average of 175,000 news articles per day in up to 73 languages and classifying them, the EMM applications apply a number of text mining and processing tools for about twenty languages. The text processing tools include news clustering, information extraction and disambiguation (persons, organisations, locations, quotations, events), matching of name variant spellings, topic detection and tracking, cross-lingual news cluster linking, opinion mining, multi-document summarisation, and more. Developing these tools is particularly challenging for highly inflected languages, such as those of the Slavic and the Finno-Ugric language families. The speaker will thus focus part of his talk on insights regarding the treatment of highly inflected languages, especially regarding information extraction and multi-label document classification. EMM is freely accessible to the public via http://emm.newsbrief.eu/overview.html.
Ron Cole (Mentor InterActive Inc. and Boulder Language Technologies, Boulder,
Spoken Dialogs with Children for Science Learning and Literacy.
Abstract: Advances in human language and character animation technologies have enabled a new generation of intelligent tutoring systems that support conversational interaction between young learners and a lifelike computer character that was designed to behave like a sensitive and effective human tutor. My Science Tutor is a spoken dialog system in which children learn to construct science explanations through conversations with Marni, the virtual science tutor, in multimedia environments. MyST displays illustrations, silent animations or interactive simulations to the student, while Marni asks open-ended questions like "What's going here?". Based on MyST's analysis of the student's spoken response, the system decides what the student understands about the science and what the student has not yet explained (or doesn't know), and generates a follow-on question a new prompt, and possibly a new animation, that is designed to scaffold learning and challenge the student to reason about the science. Two large scale evaluations were conducted in which third, fourth and fifth grade students received over 5 hours of tutoring during sixteen 20-minute sessions in four different areas of science. The results revealed that, relative to students who did not receive tutoring, students who used My Science Tutor achieved significant learning gains in standardized tests of science achievement, equivalent to gains achieved by students who received tutoring by expert human tutors. In recent research, we have extended the technologies used in MyST to develop a new generation of interactive books that use text, speech and dialog technologies to help children learn to read science texts fluently, expressively, and with good comprehension. We will demonstrate these MindStars Books and present initial results of classroom testing.
Viktor Zakharov (Saint Petersburg State University, Russia):
Russian Corpora: Comparison and Usage.
Abstract: The paper describes the existing Russian corpora and gives a brief survey of the history of corpus linguistics in Russia. The main corpora of the modern Russian language are presented, text corpora, as well as oral, parallel, and specialized ones. The corpora are compared from the point of view of their volume, search tools, and pecularities. Some examples of language investigation are given.