Conversational Interaction Conference 2019 Presentations

Monday, March 11  | KEYNOTE

Social Applications of Conversational Interfaces

Michael McTear, Professor, Ulster University

Conversational interfaces (also known as chatbots, virtual assistants, or digital assistants) have been used widely in business applications such as marketing and customer service. Increasingly attention is being directed to social and humanitarian uses of conversational interfaces, such as providing companionship and advanced home monitoring of the elderly, healthcare applications, dealing with harassment at work, providing people in developing countries with new forms of support, and many others. This talk presents some chatbot applications currently under development at Ulster University in the areas of monitoring the elderly at home, mental health support, and bibliotherapy for disadvantaged users, and examines the issues and technical challenges that these types of conversational application encounter. The talk concludes with a discussion of the current state of conversational interfaces and suggestions for future developments.

Monday, March 11  |  Tracks 1 & 2

TRACK 1

The state of conversational technology

Speech recognition and natural language understanding are developing technologies. What is their status today and how will they evolve?

TRACK 2

The voice and personality of your application

The quality, naturalness, and "personality" of speech from an automated system can be a major determinant of its effectiveness.

 

TRACK 1

Case studies 1: Experience with deployed conversational systems

Examples often make clear the advantages and challenges of a conversational system. This session contains such examples of deployed systems and their results.

TRACK 2

Advanced features in conversational systems

Conversational systems can have advanced features such as security and dealing with more than one speaker in a conversation. 

TRACK 1

Customer service and sales: Automation with natural language

Customers are frustrated with IVR menus. Improve self-service by letting them simply say or text about why they are contacting customer service.​

TRACK 2

Options in user interface design

Natural language processing technology can be poorly implemented. What are some of the considerations that make it most effective?

  • Which metaphors do and don't work for digital assistants?, Michael Greenberg, Voice User Interface Designer, Google

  • Affect, the ultimate differentiator - The Engineering of Emotion, Wolf Paulus, Principal Engineer, Intuit Futures, Intuit

  • Chat vs. Voice: The Future of Conversational AI, Arte Merritt, CEO & co-founder, Dashbot

TRACK 1

Digital assistants and chatbots: The options

What are the options in interacting with digital systems with human ("natural") language? How does one judge the tradeoffs and platforms?​

TRACK 2

Issues in creating conversational applications

Talking or text with a user of an application in a way that doesn’t require instruction or detailed prompts is clearly a potentially powerful way of interacting with an automated system. But there are challenges created by today’s state of technology.

TRACK 1

Digital Assistants in the Enterprise

Digital assistants can make employees more efficient, making the use of enterprise software and operations easier and thus more fully utilized. ​

  • Beyond Chatbots: How Augmented Intelligence Will Empower Employees to Excel, David Nahamoo, CTO, Pryon

  • Conversational EdTech for Scalable Education, Dee Kanejiya, founder and CEO, Cognii

  • Collecting Data in the Food Supply Chain using voice interaction: Updated Case Study, John Swansey, CoFounder and Chief Design Officer, AgVoice

TRACK 2

Fitting the application: Differences driven by the environment and the user

Particular cases, such as hands- and eyes-free options in automobiles or home speakers and the availability of interface options other than speech, create specialized demands on a conversational system.

TRACK 1

Robots and social systems

Science fiction is coming closer to reality as limited-function robots reach commercial feasibility. Robots and other devices that simulate human traits must not only interact naturally, but display social awareness.

  • Designing Valuable Experiences for Social Robots, Matt Willis, Lead UX Designer, SoftBank Robotics America

  • Talking to Robots: Making it Work!, Charles Jankowski, Director, AI and Robotics, Cloudminds

TRACK 2

Building conversational systems that engage users

"Conversational" interfaces raise expectations. How can we best meet those expectations?

TRACK 1

Best Practices in developing conversational systems

What practices will make a conversational system most effective?​

  • An Introduction to the Bixby Platform, John Alioto, Chief Evangelist, Viv Labs / Samsung Electronics

  • Virtual Assistants are both marvelous and god awful!, Alexandros Potamianos, CEO & co-Founder, Behavioral Signals

  • Building Actions for the Google Assistant: Lessons from the Google I/O '18 Action, Sachit Mishra, Developer Programs Engineer, Actions on Google

 

TRACK 2

Technology challenges in conversational systems

The technology supporting conversational systems has passed a “tipping point” of practical deployment, but is still evolving, and there are many challenges.

 

Tuesday, March 12 |  KEYNOTE PANEL

Building a digital assistant or chatbot--What works and what doesn't

Experienced experts will discuss the issues in building effective conversational systems. Is this a major trend or a fad? What is the status of the core technology and development tools? What are best practices? What are potential mistakes to avoid?

Patrick Dexter, COO, Cepstral; Dr. Srinivas Bangalore, Director of AI Research Technologies, Interactions; Mike de la Cruz, CBO, Directly; Diego Ventura, CEO, noHold

Tuesday, March 12  |  Tracks 1 & 2

TRACK 1

Digital assistants in customer service

Customer service is a key area where conversational systems can have an impact, getting away from a frustrating series of menus to a prompt such as “Please say why you are calling.”

TRACK 2

Options in developing a conversational assistant

Involvement in building an intelligent virtual assistant can range from simply providing guidance and examples to a solutions provider to getting into the details. What are the options and tradeoffs?

TRACK 1

Understanding your customer: A key to effective conversational systems

There are many aspects to building a conversational system. But perhaps the most fundamental is understanding what the customer may say.

TRACK 2

Natural language interaction in complex customer service

Some customer service automation using speech or text interaction is complex due to the wide variation in the way requests are posed.

TRACK 1

Making workers more efficient

Speech recognition and natural language understanding can be used to help workers do their jobs more efficiently.​

TRACK 2

Naturally! Not so easy

When a digital system uses human language - "natural language" - it creates expectations. How well can today's technology meet those expectations?

TRACK 1

Case studies 2: Experience with deployed conversational systems

Examples often make clear the advantages and challenges of a conversational system. This session contains such examples of deployed systems and their results.

TRACK 2

Natural language tools: A wide range of options

When “machine learning” is used to develop natural language applications, it can sound intimidating. This session presents insights making those technologies more accessible.

  • A data-driven approach to building AI-powered customer service virtual agents, Ofer Ronen, General Manager, Chatbase (Area 120 product at Google)

  • Deep Reinforcement Learning for Conversational AI, Dr. Sid J Reddy, Chief Scientist, Conversica

TRACK 1

Building digital assistants intelligently

The goal of automated interaction by natural language processing--text or voice—is to indeed seem natural. This session describes how to make such interaction more efficient and effective. 

TRACK 2

Challenges in conversational interaction

Conversational interaction is a young technology. Some things you should consider…​

TRACK 1

Case studies 3: Experience with deployed conversational systems

Experience with deployed conversational systems.

 

TRACK 2

Creative improvement to the user experience

Conversational technology is evolving rapidly as we learn how to make it more effective in all situations.

  • w-facebook
  • Twitter Clean