Alexander I. Rudnicky
I'm a Professor Emeritus at the Language Technologies Instutute in the School of Computer Science at Carnegie Mellon University. I'm a part of the Carnegie Mellon Speech Group, one of the oldest such groups in the country, founded by Raj Reddy. I'm also the Director of the Carnegie Mellon Speech Consortium, a partnership between Carnegie Mellon University and industry. I am part of the faculty in the Language Technologies Institute and, actually, most of my students are based there.
My current interests center on language-based communication between humans and robots and on aspects of core speech recognition, such as out-of-vocabulary (OOV) word processing. I am also interested in approaches to learning based on implicit supervision and on improvement of speech system knowledge through dialog. Check this list of publications.
Over time my research interests have revolved around speech perception and recognition, speech interfaces, spoken dialog systems and language in general. I've headed the SpeechWear project that produced an early mobile speech system; I'm still interested in the topic of mobile speech. I also headed the Communicator project, which dealt with spoken dialog system architectures. Olympus, a successor system, is available in Open Source. The Ravenclaw dialog manager came out of the Communicator work and is the current foundation for many of the systems that we create. RavenClaw incorporates dialog management ideas that were first introduced in the Office Manager (OM) system and further developed in successor systems, including Scheduler and AGENDA. Papers listed on the publications describe this work. Some videos from our group.
In 1996 I implemented a set of web-based tools that are used to generate a knowledge base for the open source Sphinx recognition system, including a language model and a pronouncing dictionary. It's proved to be quite popular, and I've continued to maintain the tool and expand its capabilities. Do try it out if you are building a recognition system (the model formats work for any ARPA-compliant system). We continue to maintain Ravenclaw and the Olympus dialog system toolkit in open source.If you'd like to know more about me read this biographical sketch or look at this maybe not up-to-date resumé. You can also find me at LinkedIn and on Plaxo; these sites will have somewhat different information about me. I'm involved in several organizations focused on spoken language interaction (SIGdial) and speech technologies (AVIOS).
The following papers cover our work in the past while. It's probably out of date. Check here for a complete list. Or just look me up here.
Chen, YN, Sun, M,. Rudnicky A.I., and Gershman, A. Leveraging Behavioral Patterns of Mobile Applications for Personalized Spoken Language Understanding, Proceedings of The 17th ACM International Conference on Multimodal Interaction (ICMI 2015), Seattle WA, (to appear).
Chen, YN, Wang, WY and Rudnicky, A.I. Learning Semantic Hierarchy with Distributional Representations for Unsupervised Spoken Language Understanding, Proceedings of The 16th Annual Conference of the International Speech Communication Association (INTERSPEECH 2015), Dresden DE, 2015.
Sun, M., Chen, YN and Rudnicky, A.I. Learning OOV through Semantic Relatedness in Spoken Dialog Systems Proceedings of The 16th Annual Conference of the International Speech Communication Association (INTERSPEECH 2015), Dresden DE, 2015.
Chiu, J, Yajie Miao, Y, Black, AW and Rudnicky, A.I. Distributed Representation-based Spoken Word Sense Induction, Proceedings of Interspeech, Dresden DE, 2015.
Marge, M. and Rudnicky, A. Miscommunication Recovery in Physically Situated Dialogue Proceedings of SIGdial, Prague CZ, 2015.
Chen, YN, Wang, WY, Gershman, A. and Rudnicky, A.I. Matrix Factorization with Knowledge Graph Propagation for Unsupervised Spoken Language Understanding, Proceedings of The 53rd Annual Meeting of the ACL and The 7th International Joint Conference on Natural Language Processing of the Asian Federation of Natural Language Processing (ACL-IJCNLP 2015), Beijing China, 2015.
Chen YN, Wang, WY, Rudnicky, A.I., Jointly Modeling Inter-Slot Relations by Random Walk on Knowledge Graphs for Unsupervised Spoken Language Understanding, Proceedings of NAACL, Denver CO, 2015.
Yu, Z., Papangelis, A. & Rudnicky, A. TickTock: Engagement Awareness in a non-Goal-Oriented Multimodal Dialogue System. AAAI Spring Symposium, 2015.
Chen, YN & Rudnicky, AI Dynamically Supporting Unexplored Domains in Conversational Interactions by Enriching Semantics with Neural Word Embeddings Proceedings of SLT, December 2014, Lake Tahoe, NV.
Chen, YN, Wang, WY & Rudnicky, AI Leveraging Frame Semantics and Distributional Semantics for Unsupervised Semantic Slot Induction in Spoken Dialogue Systems Proceedings of SLT, December 2014, Lake Tahoe, NV.
Pappu, A. & Rudnicky, A.I. Learning Situated Knowledge Bases through Dialog. Proceedings of Interspeech, September 2014, Singapore.
Justin Chiu, J., Wang, Y., Trmal, J., Povey, D., Chen, G., Rudnicky, A. Combination of FST and CN Search in Spoken Term Detection Proceedings of Interspeech, September 2014, Singapore.
Qin, L. & Rudnicky, AI Building a vocabulary self-learning speech recognition system, Proceedings of Interspeech, September 2014, Singapore.
Pappu, A. & Rudnicky, A.I. Knowledge Acquisition Strategies for Goal-Oriented Dialog Systems. Proceedings of SIGDIAL, June 2014, Philadelphia, PA.
Chen, YN & Rudnicky, AI Two-Stage Stochastic Natural Language Generation for Email Synthesis by Modeling Sender Style and Topic Structure, Proceedings of the 8th Int'l Natural Language Generation Conference (INLG), June 2014, Philadelphia, PA.
Pappu, A., Sun, M., Sridharan, S. & Rudnicky, A.I. Conversational Strategies for Robustly Managing Dialog in Public Spaces Proceedings of EACL Dialog in Motion Workshop, 2014, Gothenburg, Sweden.
If you need to reach me, E-mail is usually a good bet, particularly when I'm away on travel, since I do check my mail regularly.
office is room 6511 in the Gates-Hillman
Center (10a on the map). Or you can reach me in one of the following ways:
Prof. Alexander Rudnicky
School of Computer Science
Carnegie Mellon University
5000 Forbes Avenue
Pittsburgh, PA 15213