My Pic My Name Text
the ability to predict!!
My Name Text
email: raveesh [at] kth [dot] se

Phone: +46-(0)8-790-7872

Fax: +46-(0)8-790-7854

About me

I am a postdoctoral researcher at the Department of Speech, Music and Hearing, Kungliga Tekniska Högskolan (KTH | Royal Institute of Technology), Stockholm, Sweden. Earlier, in March 2016, I earned a PhD with specialization in Speech Communication from the same department. The title of my PhD thesis is Data-driven Methods for Spoken Dialogue Systems: Applications in Language Understanding, Turn-taking, Error Detection, and Knowledge Acquisition"

My research interests lie in research and development of the Spoken Dialogue Systems (SDS's), which can be described as application interfaces that have the potential of endowing machines (such as computers, cars, robots, smartphones) with the ability to converse with humans in naturally spoken language. Yes, in spoken English or Swedish! Examples from everyday use of dialogue system is the Apple's speech enabled personal assistant Siri and the voice-operated automatic telephony systems. Robots with conversational skills are gradually making their way into everyday human lives, e.g., the Furhat social robot.

While dialogue systems technologies has been around for over three decades now spoken interactions with many of the aforementioned systems is far from being natural and convenient. The idea of a "conversational computer" hence remains an open challenge for researchers in artificial intelligence community. In my research work I aim at how machine learning methods can be used to build more robust dialogue systems. In my PhD thesis I have investigated the use of data-driven methodology for tasks such as language understanding, turn-taking, error detection, and knowledge acquisition.

As part of my postdoctoral research I am investigating how to build dialogue systems that can sustain meaningful long term interactions, i.e., systems that can be used as companions to whom users can talk on everyday basis. An important aspect for the system here is how to sustain such an interaction.

Research Interests

Spoken language dialogue systems, social robots, conversational analysis, user evaluations, machine learning techniques for dialogue system, dialogue modelling.

General Interests

Machine learning and algorithms, better programming practices (Java), OpenSource Development.

Discussions on sociocultural evolution, history, science and society, beer, cooking, outdoor actvity (kayaking!).

If you share similar interests please feel free to get in touch with me. I'm always eager to have discussions on the hows and the whats of research objectives. Drop me a mail.

Publications

To my university profile-page.

Korbayová, I., Meena, R., Pyykönen, P. (2011). Perception of Visual Context and Intonation Patterns in Robot Utterances. In proceedings of 6th ACM/IEEE International Conference on Human-Robot Interaction (pp. 173-174). Lausanne, Switzerland. [pdf]

Research Training

# The 8th International Summer Workshop on Multimodal Interfaces , Metz, France. July 2-27, 2012. I was involved in the project "Speech, gaze and gesturing - multimodal conversational interaction with Nao robot."

# Research Training Course: Feedback, Communicative Gesturing, and Gazing, Helsinki, Finland. May 23-28, 2011. Here is a report from the training course Hands-on Experience on Conducting an Experimental Study on Engagement in Human-Human Conversation. Raveesh Meena, Angelo Cafaro, Hana Vrzáková, Andreas Falck, Max Bohlom, Gustaf Lindblad. [pdf].

Activities

1. Member of Organising Committee AWoCoDo'12 - the Annual Workshop for Computer Science Doctoral Students at CSC/KTH.

2. Member of Organising Committee YRRSDS'12 & YRRSDS'14 (Young Researchers Roundtable on Spoken Dialogue Systems)

3. Member of Postgraduate Students' Council at KTH/CSC (2011-2014). Here I worked in the role of (a.) PhD student representative from the TMH department, (b.) PhD student representative from TMH in the CS Doctoral Program Committee, and (c.) Member of Employment Committee.

4. I used to play Ultimate (Frisbee) with Stockholm Syndrom.

5. I like beign outdoors, paddeling, cykling, and hiking on the weekends

# During my I was also a PhD-associate in the Swedish National Graduate School of Language Technology (GSLT).

Previous work

I got practically involved with research on dialogue systems at the Talking Robots lab while working as a research assistant during my M.Sc studies. My master thesis focuses on the "presentational" aspects of system generated utterances, particularly clarification requests and information seeking questions. In my master thesis, I developed a novel approach for the production of contextually appropriate intonation in a robot's clarification requests. The ability of a robot in explicitly controlling the intonation of its utterances enables it to convey the intended meanings behind the two utterances: "Is that a RED mug?" vs. "Is that a red MUG?" (where the words in captial have the nuclear stress is the sentence).

M.Sc. thesis: Contextually Appropriate Intonation of Clarification Requests in Situated Human-Robot Dialogue. Angefertigt unter Anleitung von Dr.ing. Ivana-Kruijff-Korbayová und Prof. Dr. Hans Uszkoreit. July 2010. Saarland University, Germany

Brief biography

I come from India (Rajasthan). I grew up studying in various north-west and central cities in India. I earned my Bachelors in Information Technology (1999-2003) from the Indian Institute of Information Technology, Allahabad, (India). This is where I first came in touch with the field of Natural Language Processing (NLP). During one of the project works I, together with two other classmates, studied various machine learning approaches for document classification (we used e-mail classification as the domain for implementation). In another project, we used string-matching algorithms (taught in a course on Bioinformatics) for detecting plagiarism in student essays.

After my Bachelor's, I worked for almost 4 years as a software engineer (2004-2007) for the Tata Counsultancy Services Ltd. (TCS), India. There I worked with the architecture team and co-lead development and ehnacement of the core architecture of a TCS financial services product, namely QUARTZ, which is now known as BaNCS in the international BFSI market. During this time I also co-developed and deployed a template based alert and notification engine that sent out messages to intended recipients. For example, a pop-up message to the fund manager for verifying a deal, or a SMS notification to the bank customers about transactions in their accounts.

In 2006, I was briefly on deputation in the Netherlands. I mainly helped in the deployment of the alert and notification engine, and assisted in performance tuning of the various algorithms and database queries in a production environment.

During this brief stay in Europe I was highly impressed with the use of technology in everyday life. A wishful thought of making such technology accessible to the (illiterate) masses in India motivated me to think of speech enabled interfaces. This wishful thinking was instrumental in my application to the International M.Sc. program in Language Science and Technology at the Department of Computational Linguistics and Phonetics, Saarland University, Germany. During my master studies I developed special interest in spoken dialogue systems (SDS). Alongside my studies, I worked as a Research Assistant in the Talking Robots group, which was part of the Language Technology Lab at the German Research Center for Aritificial Intelligence, (DFKI GmbH), Saarbrücken. I pursued my master thesis in this group. I finished my M.Sc in July 2010, and since Feb 2011 I'm at KTH.

Disclaimer

Disclaimer for the personal web pages