[PYTHON] The story of creating a store search BOT (AI LINE BOT) for Go To EAT in Chiba Prefecture (2) [Overview]
This is a continuation of the story (1) of creating a store search BOT (AI LINE BOT) for Go To EAT in Chiba Prefecture.
The system configuration is as follows.
【flow】
- Operate the LINE BOT on your smartphone and enter the word related to the store you want to search.
- When you enter it, an https request will be sent to the WEB server on the AI search server via LINE's Messageing API.
- Requests are accepted by flask, and shops with high similarity to the search word are identified by AI.
- Based on the result determined by AI, we will return the information about the store from the database.
To avoid spending money, the AI search server adopted t2.micro, which is provided as a free tier on AWS. However, since the memory is small to create an AI model, I used the surface pro 6 at hand for the model creation work itself.
To create the AI model, we use the information on the tabelog site and the list of shops on the Go To eat Chiba Prefecture website (shop_list.pdf) (the list of shops is currently unavailable for download).
In the following articles, we will talk about the construction procedure.