Astrodetection
Astrodetection is a Python library designed for detecting astroturfing clues from lists of posts (mainly on X up to now, but not exclusively)
Installation
Pip
pip install "astrodetection[standard]"
or
pip install "astrodetection[light]"
Conda
-
Use the YAML file to configure the environment with conda:
conda create -n astrodetection_env conda activate astrodetection_env conda env update -f environment_standard.yml
Note: the environment_standard.yml configuration file uses FAISS and Fasttext libraries for VIGINUM D3LTA implementation
**If you have compatibility issues, prefer environment_light.yml and use astrodetection_light module
Usage
You can import directly the main functions:
from astrodetection import semantic_faiss, prepare_input_data, compute_bot_likelihood_metrics, create_network
Or use them directly:
import glob
import pandas as pd
import os
import numpy as np
import astrodetection
# Load a single JSON file into a DataFrame
file = "file_path" # Select the first file
df = pd.read_json(file)
df.index = df.index.astype(str) # Compatibility with d3lta
# Preprocess the DataFrame
df = df[df['tweet'].str.len() > 100]
df = df[df['username'] != 'grok']
df.index = df.index.astype(str)
# Compute matches and scores
df_filtered, df_emb = astrodetection.prepare_input_data(df, embeddings=df['emb'])
matches, df_cluster = astrodetection.semantic_faiss(
df_filtered.rename(columns={'tweet': 'original'}),
min_size_txt=0,
df_embeddings_use=df_emb,
threshold_grapheme=0.8,
threshold_language=0.715,
threshold_semantic=0.9
) #function taken from D3LTA
scores = astrodetection.compute_bot_likelihood_metrics(df, matches=matches)
# Create a network
network = astrodetection.create_network(matches, df)
New changes
-
semantic_faissfunction can now take detect only copypastas based on levenshtein distance, ignoring embeddings, if "skip" is passed as argument in df_embeddings_use field. -
compute_bot_likelihood_metricsfunction can now take columns' names as arguments for more customization -
compute_bot_likelihood_metricsnow returnshigh_following_followers_ratio (%), the share of rows whose following/followers ratio exceeds fw_ratio_threshold (default 10).
1 maintainer
Release files for astrodetection 0.3.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| astrodetection-0.3.0.tar.gz | 55.8 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| astrodetection-0.3.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 113.0 kB
Release files / astrodetection-0.3.0.tar.gz
| Download URL | astrodetection-0.3.0.tar.gz |
|---|---|
| Size | 55.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
3d048129f5a1f302bc070d721562924dd1d400759768413c39a0116971ab01f2
|
|
BLAKE2b-256 checksum How to use checksums |
97929e052d4e81a80cf6f925828db1f5efc7df575aa44a7bc2a4dae5bfe4aba5
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.4
|
Release files / astrodetection-0.3.0-py3-none-any.whl
| Download URL | astrodetection-0.3.0-py3-none-any.whl |
|---|---|
| Size | 57.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
85a302f53453f4cce9083ca872763a09f143386a86183e7ff8afe8b0ad4627ee
|
|
BLAKE2b-256 checksum How to use checksums |
26995f7f2966f031432359fee4ae0e00e9f614ea94839c2901fcd1e016ff4053
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.4
|
