Autor: |
Shrimai Prabhumoye, Ryan Lowe, Kurt Shuster, Joelle Pineau, Alexander H. Miller, Emily Dinan, Douwe Kiela, Jack Urbanek, Iulian Vlad Serban, Valentin Malykh, Arthur Szlam, Alexander I. Rudnicky, Varvara Logacheva, Mikhail S. Burtsev, Jason D. Williams, Jason Weston, Alan W. Black |
Rok vydání: |
2019 |
Předmět: |
|
Zdroj: |
The NeurIPS '18 Competition ISBN: 9783030291341 |
DOI: |
10.1007/978-3-030-29135-8_7 |
Popis: |
We describe the setting and results of the ConvAI2 NeurIPS competition that aims to further the state-of-the-art in open-domain chatbots. Some key takeaways from the competition are: (1) pretrained Transformer variants are currently the best performing models on this task, (2) but to improve performance on multi-turn conversations with humans, future systems must go beyond single word metrics like perplexity to measure the performance across sequences of utterances (conversations)—in terms of repetition, consistency and balance of dialogue acts (e.g. how many questions asked vs. answered). |
Databáze: |
OpenAIRE |
Externí odkaz: |
|