rgmining-tripadvisor-dataset

Trip Advisor dataset for Review Graph Mining Project


Keywords
dataset, graph, mining, review
License
GPL-3.0
Install
pip install rgmining-tripadvisor-dataset==0.5.6

Documentation

Trip Advisor Dataset Loader

GPLv3 Build Status wercker status Release PyPi Japanese

Logo

For the Review Graph Mining project, this package provides a loader of the Trip Advisor dataset provided by Dr. Wang.

Installation

Use pip to install this package.

$ pip install --upgrade rgmining-tripadvisor-dataset

Note that this installation will download a big data file from the original web site.

This package uses bz2 internally. If your python doesn't have that package (try import bz2), rebuild python before installation.

Usage

This package provides module tripadvisor and this module provides load function. The load function takes a graph object which implements the graph interface defined in Review Graph Mining project.

For example, the following code constructs a graph object provides the FRAUDAR algorithm, loads the Trip Advisor dataset, runs the algorithm, and then outputs names of anomalous reviewers. Since this dataset consists of huge reviews, loading may take long time.

import fraudar
import tripadvisor

# Construct a graph and load the dataset.
graph = fraudar.ReviewGraph()
tripadvisor.load(graph)

# Run the analyzing algorithm.
graph.update()

# Print names of reviewers who are judged as anomalous.
for r in graph.reviewers:
  if r.anomalous_score == 1:
    print r.name

# The number of reviewers the dataset has: -> 1169456.
len(graph.reviewers)

# The number of reviewers judged as anomalous: -> 147.
len([r for r in graph.reviewers if r.anomalous_score == 1])

Note that you may need to install the FRAUDAR algorithm for the Review Mining Project by pip install rgmining-fraudar.

License

This software is released under The GNU General Public License Version 3, see COPYING for more detail.

The authors of the Trip Advisor dataset, which this software imports, requires to cite the following papers when you publish research papers using this package: