search

Conchylicultor / DeepQA

My tensorflow implementation of "A neural conversational model", a Deep learning based chatbot
Apache License 2.0
2.93k stars 1.17k forks source link

Question Regarding the Community QA or Long Answers #106

Open nick-magnini opened 7 years ago

nick-magnini commented 7 years ago

Hi,

Based on some improvements mentioned in the description, how can DeepQA be adapted to community QA learning? As an example: Q: SentenceA. --> A: Sentence1. Sentence2. Sentence3. Sentence4. Sentence5. Would it be fine if this is modeled as Q: SentenceA. --> A: Sentence1. Sentence2. Sentence3. Sentence4. Sentence5. or better to be modeled as oversampled examples in this way: Q: SentenceA. --> A: Sentence1. Sentence2. Q: SentenceA. --> A: Sentence2. Sentence3. Q: SentenceA. --> A: Sentence3. Sentence4. Q: SentenceA. --> A: Sentence4. Sentence5. Q: SentenceA. --> A: Sentence2. Sentence3. Sentence4. ...

Thanks.