Blockchain Social Media

Requirements

Python version: 3.10.12

Some pacakges require apt or conda to be installed.

environment.yml requirements.txt

To update the requirements file, run

pip list --format=freeze > requirements.txt

To update the conda environment file, run

conda env export > environment.yml

Tweet ids:

The tweet ids are available under tweet_ids/*.json

How to use:

.env file setup

The .env file must be populated with the following variables:

ES_HOSTNAME={ADDRESS_TO_ELASTICSEARCH}
DATA_DIR={ABSOLUTE_PATH_TO_DATA_DIR}

The code will use the .env file to load the environment variables.

Elasticsearch setup

The settings must be set in the config files. An example config can be found in ./config/twitter:blockchain-interoperability-attacks

Loading the data

The raw snapshots are stored as pickle files. The CryptoChatterData object will store individual columns separately after parsing. This allows the data to use only related columns.

from crypto_chatter.data import CryptoChatterData
from crypto_chatter.graph import CryptoChatterGraph
from crypto_chatter.config import (
    CryptoChatterDataConfig,
    CryptoChatterGraphConfig
)

dataset = "twitter:blockchain-interoperability-attacks"
data_config = CryptoChatterDataConfig(dataset)
data = CryptoChatterData(
    data_config,
)

data.load([data.data_config.clean_text_col])

graph_kind = f"tweet-reply"
graph_config = CryptoChatterGraphConfig(
    data_config=data_config,
    graph_kind=graph_kind,
)

builder = CryptoChatterGraphBuilder(
    data=data,
    graph_config=graph_config,
)

graph = builder.get_graph()