Hi @stephens Is it possible to extract my conversations data in csv or excel format. If yes how. I have installed rasa x using docker-compose mode. I want to extract my data beaucause i want to do some analysis. Thanks.
Yes, you could do this with Postgres tools to pull it from the Rasa X db. There’s a dump tool with postgres that I’ve used.
However, unless this is a one time analysis, the best mechanism is to send the Rasa OSS event broker stream to an additional message queue. With both Rabbit and Kafka, you can define a second message queue and consume that with almost any analytics tool. I’ve used the Elastic stack, reading the message queue with logstash.
Thank you for replying but i want to know which dump tool did you use?
@Papa see the concept of conversation= tracker.events in action.py and try the python script to save the data in .csv/xls or txt. But when you save the data it will save in docker container so you need to used docker cp <container id>:"/app" "."
@stephens I had even tried the Elastic (ELK) Stack to create a dashboard in Kibana and I successfully developed one.But, one drawback I faced, the log conversation data when we stored in PostgresDB the parsed data column is in JSON(metadata; which have all the main record of logs), so we first need to change to text and then for indexing and at the backend it’s very complex process. But it was fun 