Semantic Similarity Chatbot

The EMOTIVE project is using chatbots to foster archaeological empathy. I’m wondering, can I use a chatbot as a way of exploring the latent idea space of published archeological knowledge?

Allison Parrish’s bot demonstrates how to build a chatbot trained on the Cornell Movie Dialogue Database. My ambition here is to train the bot on the written outputs of the CRANE project: an informed archaeological chatbot. The chatbot learns the archaeological words and via word embeddings the universe of related concepts.

The hard part is going to be to get the training data into a structure that makes sense for a conversation. This might be a fool’s errand.

View and run Allison Parrish’s chatbot on Google Colab or in

Parrish, Allison. 2018. Semantic Similarity Chatbot

Roussou, M., Perry, S., Katifori, A., Vassos, S., Tzouganatou, A., McKinney, S. (2019) Transformation through Provocation? Designing a ‘Bot of Conviction’ to Challenge Conceptions and Evoke Critical Reflection. In CHI ’19 Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems, Glasgow, Scotland, 4-9 May. New York: ACM. Paper No. 627.

Other Possible Models

How I Used Deep Learning To Train A Chatbot To Talk Like Me (Sorta) | github

List of chatbot frameworks, code, etc

A chatbot trained on reddit ~ perhaps extract the archaeology reddit threads…

DeepQA ~ has examples of training on your own dataset

…to be continued…