All posts by Todd

BostonCHI hosts Cody Dunne at Audible: Interactive network and time series visualizations for reasoning and communicating about data

Cody Dunne

Cody Dunne, PhD, Assistant Professor, Northeastern University

Tuesday, May 8, 2018 at 6:30 PM

At Audible, 101 Main Street, 6th Floor, Cambridge, MA 02142

Please register, registration is required. Note that you’ll be required to sign an Amazon NDA and show an ID.

Abstract

The modern world is awash in complex data that can contain the keys to improving our lives. The scope of this data has rapidly outpaced our capabilities to analyze and comprehend, so we turn to computers to help. However, state-of-the-art technology can only supplement the human element. People assist in each stage of data science, whether it’s data cleaning, understanding algorithm design, exploring computed results, or collaborating and sharing for decision-making. To present complex information to humans, we use visualizations that leverage our extraordinary perceptual system which can detect trends, clusters, gaps, and outliers almost instantly.

A challenging and increasingly important type of data is networks of entities and their relationships. Networks are widely used across diverse disciplines to reason about complex behavior. These analyses involve understanding relationships, as well as associated attributes, statistics, or groupings. The omnipresent node-link visualization excels at showing topology and features simultaneously, but many are difficult to extract meaning from due to poor layout or shoehorning inherent complexity into limited space. Some networks have geospatial components and are temporally evolving – which further complicates visualization design. The first part of my talk will detail techniques for measuring the readability of node-link visualizations and strategies to help users create more effective and understandable visualizations in several domains.

Temporal datasets may also have interrelated variables that can make temporal inference difficult. As the volume and frequency of data grows it becomes difficult to see the subsets of data relevant to an event, or set of events, of significance to the domain problem. These issues are compounded when there is poor data quality such as missing data and uncertainty in values or timestamps. The second part of my talk will describe ongoing research in my group on methodologies for visualizing long streams of temporal event sequences with multidimensional, interrelated data. I will present a case study on using these techniques for type 1 diabetes clinical decision support.

Bio

Professor Cody Dunne works at the intersection of information visualization, network science, human-computer interaction, and computer science. He focuses on techniques for making data easier to analyze and share, as well as the application of visualization techniques to real-world problems. Dr. Dunne is currently researching the next generation of techniques for visually exploring, sharing, and collaborating around data.

Some domains Dr. Dunne has worked on include visualizing events and concepts from medical records, the spread of infectious diseases, citations in academic literature, interactions of people and organizations, relationships in archaeological dig sites, news term co-occurrence, thesaurus category relationships, municipal energy use, and computer network traffic flow.

Prior to joining Northeastern in 2016, Dr. Dunne was a research scientist in IBM Watson Health, IBM Watson, and IBM Research. Dr. Dunne received his PhD and MS degrees in computer science under Ben Shneiderman at the University of Maryland Human-Computer Interaction Lab in 2013 and 2009, respectively. He earned a B.A. degree in computer science and mathematics from Cornell College in 2007.

Evening schedule

6:30 – 7:00  Networking over pizza and beverages
7:00 – 8:30  Meeting
8:30 – 9:00  CHI Dessert and more networking

Sponsors

Thank you to our generous sponsors. Interested in sponsoring BostonCHI? Let us know!

Audible is hosting.

Izotope is sponsoring dessert.

Getting there

Audible is located at 101 Main Street, 6th Floor, Cambridge, MA 02142. The entrance to the building is right next to Tatte Bakery on Main Street. The best way to get there is to take the Red Line to Kendall Square. If you’re driving, here’s a list of parking locations.

.

 

Donna K Byron, IBM Watson & Cloud Platform, Good morning, beautiful chatbot! 5 best practices for rapid development of effective conversational agents

 

Donna K. Byron, PhD

IBM Watson and Cloud Platform division, Littleton MA

Good morning, beautiful chatbot! 5 best practices for rapid development of effective conversational agents

Tuesday, April 10, 2017 at 6:30 PM

At Audible, 101 Main Street, 6th Floor, Cambridge, MA 02142

Please register, registration is required.

Abstract

The recent enthusiasm to create conversational interfaces and chatbots for e-commerce has been enabled by the availability of platforms with simple tooling for creating dialog flows, out-of-the-box language understanding, and one-click deployment to channels such as Slack. These tools bring chatbot creation within reach for makers from many professional backgrounds. While bot platforms are simple to use, many people are disappointed with their first attempts at building engaging and fluid conversations with out-of-the-box tools. But by incorporating a handful of key findings from human interaction models and spoken dialogue systems research, the linguistic and task competence of a conversational agent can be greatly improved.

This talk covers a focused set of design methods for human-computer dialogue that should be in every designer’s toolkit to field effective and engaging conversational agents, even on an aggressive timeline. They can be utilized when creating conversational systems from scratch or within a platform such as Microsoft Bot Framework, wit.ai, etc.  Dr. Byron will share how these methods have been used in the context of fielding multiple conversational systems using the Watson Conversation Service by IBM.

Bio

Donna Byron currently works as a Senior Software Developer in the Watson and Cloud Platform division at IBM. She has been building intelligent conversational agents in both commercial and research settings for 20 years.  Since joining IBM’s Watson Software group in 2012, she has built natural language understanding components in diverse domains from healthcare to cooking to pop culture and travel.

Prior to joining IBM, Dr. Byron contributed to human-computer dialogue research in several academic labs, including the Relational Agents lab at Northeastern University.  She completed her PhD in the Conversational Interaction and Spoken Dialogue Research Group at the University of Rochester, where she helped advance the state of the art in deep semantic processing for spoken dialogue agents performing collaborative tasks with human partners. She also co-founded the Speech and Language Technologies lab in the Ohio State University Computer Science Department, where her research focused on human-computer dialogue for embodied conversation in virtual environments.

Dr. Byron is the inventor or co-inventor on over 40 US Patents in various areas of Artificial Intelligence and was awarded the designation of Master Inventor at IBM in 2014.

Evening schedule

6:30 – 7:00  Networking over pizza and beverages
7:00 – 8:30  Meeting
8:30 – 9:00  CHI Dessert and more networking

Sponsors

Thank you to our generous sponsors. Interested in sponsoring BostonCHI? Let us know!

Audible is hosting us

Mad*Pow is sponsoring pizza.

Vitamin T is sponsoring dessert.

Getting there

Audible is located at 101 Main Street, 6th Floor, Cambridge, MA 02142. The entrance to the building is right next to Tatte Bakery on Main Street. The best way to get there is to take the Red Line to Kendall Square. If you’re driving, here’s a list of parking locations.

 

RESCHEDULED – Interactive network and time series visualizations for reasoning and communicating about data

 

This talk has been rescheduled to

 

 

Cody DunneCody Dunne, PhD, Assistant Professor, Northeastern University

Tuesday, March 13, 2017 at 6:30 PM

At Audible, 101 Main Street, 6th Floor, Cambridge, MA 02142

Please register, registration is required.

Abstract

The modern world is awash in complex data that can contain the keys to improving our lives. The scope of this data has rapidly outpaced our capabilities to analyze and comprehend, so we turn to computers to help. However, state-of-the-art technology can only supplement the human element. People assist in each stage of data science, whether it’s data cleaning, understanding algorithm design, exploring computed results, or collaborating and sharing for decision-making. To present complex information to humans, we use visualizations that leverage our extraordinary perceptual system which can detect trends, clusters, gaps, and outliers almost instantly.

A challenging and increasingly important type of data is networks of entities and their relationships. Networks are widely used across diverse disciplines to reason about complex behavior. These analyses involve understanding relationships, as well as associated attributes, statistics, or groupings. The omnipresent node-link visualization excels at showing topology and features simultaneously, but many are difficult to extract meaning from due to poor layout or shoehorning inherent complexity into limited space. Some networks have geospatial components and are temporally evolving – which further complicates visualization design. The first part of my talk will detail techniques for measuring the readability of node-link visualizations and strategies to help users create more effective and understandable visualizations in several domains.

Temporal datasets may also have interrelated variables that can make temporal inference difficult. As the volume and frequency of data grows it becomes difficult to see the subsets of data relevant to an event, or set of events, of significance to the domain problem. These issues are compounded when there is poor data quality such as missing data and uncertainty in values or timestamps. The second part of my talk will describe ongoing research in my group on methodologies for visualizing long streams of temporal event sequences with multidimensional, interrelated data. I will present a case study on using these techniques for type 1 diabetes clinical decision support.

Bio

Professor Cody Dunne works at the intersection of information visualization, network science, human-computer interaction, and computer science. He focuses on techniques for making data easier to analyze and share, as well as the application of visualization techniques to real-world problems. Dr. Dunne is currently researching the next generation of techniques for visually exploring, sharing, and collaborating around data.

Some domains Dr. Dunne has worked on include visualizing events and concepts from medical records, the spread of infectious diseases, citations in academic literature, interactions of people and organizations, relationships in archaeological dig sites, news term co-occurrence, thesaurus category relationships, municipal energy use, and computer network traffic flow.

Prior to joining Northeastern in 2016, Dr. Dunne was a research scientist in IBM Watson Health, IBM Watson, and IBM Research. Dr. Dunne received his PhD and MS degrees in computer science under Ben Shneiderman at the University of Maryland Human-Computer Interaction Lab in 2013 and 2009, respectively. He earned a B.A. degree in computer science and mathematics from Cornell College in 2007.

Evening schedule

6:30 – 7:00  Networking over pizza and beverages
7:00 – 8:30  Meeting
8:30 – 9:00  CHI Dessert and more networking

Sponsors

Thank you to our generous sponsors. Interested in sponsoring BostonCHI? Let us know!

Audible is hosting us and sponsoring pizza.

Izotope is sponsoring dessert.

Getting there

Audible is located at 101 Main Street, 6th Floor, Cambridge, MA 02142. The entrance to the building is right next to Tatte Bakery on Main Street. The best way to get there is to take the Red Line to Kendall Square. If you’re driving, here’s a list of parking locations.

.