[NLG]Do Neural Dialog Systems Use the Conversation History Effectively? An Empirical Study?
摘要： 爲了發現對話歷史是否真的被有效的利用了，本文介紹了10種干擾來評測對話歷史是否真正被利用。發現transformer-seq2seq 和rnn seq2seq並沒有很好地利用這些信息 主要工作： 首先有一個前提：如果模型對於對話歷史的改變不敏感的話，那麼他基本上沒有利用多少對話歷史。 在之前的工作中已經發現： 1.lstm的作用範圍只被周圍的大約150個詞所影響 2.
相關文章
- 1. An Empirical Study of Real-world Polymorphic Code Injection Attacks
- 2. An Empirical Study of Detection-Based Video Instance Segmentation
- 3. 【P2】On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study
- 4. 論文閱讀：RUBER: An Unsupervised Method for Automatic Evaluation of Open-Domain Dialog Systems
- 5. AI systems that'll help study the behavior of animal species
- 6. 論文閱讀：An Empirical Study of Spatial Attention Mechanisms in Deep Networks
- 7. A Knowledge-Grounded Neural Conversation Model 翻譯
- 8. 二值網絡訓練--A Empirical Study of Binary Neural Networks' Optimisation
- 9. [dialog system]閱讀筆記 - 2016__A Persona-Based Neural Conversation Model
- 10. 【文獻閱讀】An Empirical Study of Spatial Attention Mechanisms in Deep Networks
- 更多相關文章...
- • Docker history 命令 - Docker命令大全
- • PHP 7 use 語句 - PHP 7 新特性
- • RxJava操作符（一）Creating Observables
- • 使用Rxjava計算圓周率
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. An Empirical Study of Real-world Polymorphic Code Injection Attacks
- 2. An Empirical Study of Detection-Based Video Instance Segmentation
- 3. 【P2】On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study
- 4. 論文閱讀：RUBER: An Unsupervised Method for Automatic Evaluation of Open-Domain Dialog Systems
- 5. AI systems that'll help study the behavior of animal species
- 6. 論文閱讀：An Empirical Study of Spatial Attention Mechanisms in Deep Networks
- 7. A Knowledge-Grounded Neural Conversation Model 翻譯
- 8. 二值網絡訓練--A Empirical Study of Binary Neural Networks' Optimisation
- 9. [dialog system]閱讀筆記 - 2016__A Persona-Based Neural Conversation Model
- 10. 【文獻閱讀】An Empirical Study of Spatial Attention Mechanisms in Deep Networks