Second place was awarded to Alquist from the Czech Technical University in Prague and third place to Alana from the Heriot-Watt University in Edinburgh. Amazon wants to thank all of the Alexa Prize teams for their creativity, passion, and contributions to both AI science and delighting Alexa customers.
Winner: Gunrock
With an average score of 3.1 and average duration of 9 minutes and 59 seconds, Team Gunrock will share the $500,000 prize among the student team members.
Second: Alquist
Team Alquist will share a $100,000 prize among the student team members.
Third: Alana
Team What's Up Bot will share a $50,000 prize among the student team members.
To learn more about the detailed evaluation criteria, please read our announcement.
Proceedings
The Alexa Prize Proceedings publishes the research in Conversational AI resulting from the pursuit of the Alexa Prize competition goals.
Amazon works closely with university teams to provide a testbed for research to address the challenges with Dialog Management, Natural Language Understanding (NLU), Contextual Modeling, Commonsense Reasoning and Response Generation, and these proceedings seek to capture the advances in those areas that result from these efforts. Authors are free to make additional hardcopy publishing arrangements, but Amazon will not produce hardcopies of these volumes.
Forward
The 2018 Alexa Prize: Several Steps Forward in Conversational AI
Introduction
Introduction to 2018 Alexa Prize Proceedings
Amazon Alexa Prize
Advancing the State of the Art in Open Domain Dialog Systems through the Alexa Prize
University of California Davis – Gunrock
Building A Human-Like Social Bot By Leveraging Large Scale Real User Data
Czech Technical University in Prague – Alquist
Alexa Prize Socialbot Based on Sub-Dialogue Models
Heriot-Watt University – Alana
Entertaining and Informative Open-domain Social Dialogue using Ontologies and Entity Linking
Brigham Young University - Eve
Mixed Initiative Dialog via Structured Knowledge Graph Traversal and Conversational Scaffolding
Carnegie Mellon University - Tartan
A Retrieval-based Socialbot Powered by a Dynamic Finite-state Machine Architecture
Emory University - IrisBot
An Open-Domain Conversational Bot for Personalized Information Access
KTH Royal Institute of Technology - Fantom
A Crowdsourced Social Chatbot using an Evolving Dialog Graph
University of California Santa Cruz – SlugBot
Developing a Computational Model and Framework of a Novel Dialogue Genre
You can also download all of the papers in one .zip file.
Externally published papers
2018 Alexa Prize team papers:
- Towards Coherent and Engaging Spoken Dialog Response Generation Using Automatic Conversation Evaluators
- Contextual Language Model Adaptation for Conversational Agents
- Contextual Topic Modeling For Dialog Systems
- On Evaluating and Comparing Open Domain Dialog Systems
- Detecting Offensive Content in Open-domain Conversations using Two Stage Semi-supervision
- Topic-based Evaluation for Conversational Bots
Competing team papers:
- Contextual Dialogue Act Classification for Open-Domain Conversational Agents
- Combining Search with Structured Data to Create a More Engaging User Experience in Open Domain Dialogue
- An Ensemble Model with Ranking for Social Dialogue
- Neural Response Ranking for Social Conversation: A Data-Efficient Approach
- Automatically Predicting User Ratings for Conversational Systems
- Hybrid Code Networks using a convolutional neural network as an input layer achieves higher turn accuracy
- Crowdsourcing a self-evolving dialog graph
- SlugNERDS: A Named Entity Recognition Tool for Open Domain Dialogue Systems
- Combining Search with Structured Data to Create a More Engaging User Experience in Open Domain Dialogue
- Dependency Parsing for Spoken Dialog Systems
