/topic-model-mas

models topics via the autonolas Multi Agent System framework

Primary LanguagePythonApache License 2.0Apache-2.0

Topic-modelling AEA

Introduction

This is a topic-modelling AEA that collects textual data from third party (external) data-source and lets the agents publish the agreed-upon top 3 topics from daily updates into a shared note-taking space.

Refs:

Creating an Agent

Overview docs:

Setup

refs:

Install pre-requisites:

touch Pipfile && pipenv --python 3.10
pipenv install open-autonomy
pipenv shell

Initialize new agent

autonomy init --remote
cd packages
aea create topic_aggregator --remote --empty
aea search --local skills
# Searching for ""...
# No skills found.

Add a skill

aea scaffold skill topic_extractor

Fingerprinting packages and skills

aea fingerprint by-path packages/topic_aggregator/skills/topic_extractor 

refs:

Search local packages

aea search --local skills
Searching for ""...
Skills found:

------------------------------
Public ID: valory/abstract_round_abci:0.1.0
Name: abstract_round_abci
Description: abstract round-based ABCI application
Author: valory
Version: 0.1.0
------------------------------
------------------------------
Public ID: arco/topic_extractor:0.1.0
Name: topic_extractor
Description: The scaffold skill is a scaffold for your own skill implementation.
Author: arco
Version: 0.1.0
------------------------------

External data-source

refs:

Testing

refs:

pipenv run pytest test.py

Frontend

You could drump up a frontend via React.

refs: