Skip to main content

Automatic analysis of Airbnb host profiles.

Project description

A python package for automatic analysis of Airbnb host profiles.

The package takes an Airbnb profile, automatically tags topics for each sentence, and predicts whether the profile will be perceived as more trustworthy compared to other profiles of similar length.

Example Usage

LIWC2007 is a proprietary dependency required for predicting trust. Unfortunately, we can’t include it in this package for legal reasons.

However, if you do not have LIWC, you can still perform topic classification.

from AirProfile import AirProfile

ap = AirProfile(liwc_path='../LIWC2007/liwc_2007.trie')
# or ap = AirProfile() if you do not have LIWC.

# Example Airbnb host profile.
input = """I have spent my life in the service industry. I look forward to being your host and I look forward to meeting you."""

# Segments the input at the sentence level and returns the probability that
# each sentence is tagged with the topics described in [1]. This works with or
# without LIWC.
ap.predict_topics(input)

>>> [
  [
    'i have spent my life in the service industry',
    {
      'relationships': 0.02,
      'workEducation': 0.99,
      'travel': 0.0,
      'originResidence': 0.07,
      'lifeMottoValues': 0.03,
      'hospitality': 0.02,
      'interestsTastes': 0.03,
      'personality': 0.02
    }
  ], [
    'i look forward to being your host and i look forward to meeting you',
    {
      'relationships': 0.0,
      'workEducation': 0.0,
      'travel': 0.02,
      'originResidence': 0.0,
      'lifeMottoValues': 0.0,
      'hospitality': 1.0,
      'interestsTastes': 0.0,
      'personality': 0.04
    }
  ]
]

# Segments the input at the sentence level and returns the probability that
# the profile is perceived to be more trustworthy compared to other profiles
# of similar length. This requires LIWC and will throw an error otherwise.
ap.predict_trust(input)

>>> Prediction(prob=0.49, predict=0)

References

[1] Self-disclosure and Perceived Trustworthiness of Airbnb Host Profiles. Xiao Ma, Jeff Hancock, Kenneth Lim Mingjie, and Mor Naaman. CSCW 2017. Honorable Mention for Best Paper. [PDF1]

[2] A Computational Approach to Perceived Trustworthiness of Airbnb Host Profiles. Xiao Ma, Trishala Neeraj, Mor Naamann. ICWSM 2017. Poster. [PDF2]

Project details


Release history Release notifications

This version
History Node

1.0.12

History Node

1.0.11

History Node

1.0.10

History Node

1.0.9

History Node

1.0.8

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
AirProfile-1.0.12-py2-none-any.whl (4.3 kB) Copy SHA256 hash SHA256 Wheel py2 Sep 25, 2017
AirProfile-1.0.12.tar.gz (86.0 kB) Copy SHA256 hash SHA256 Source None Sep 25, 2017

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging CloudAMQP CloudAMQP RabbitMQ AWS AWS Cloud computing Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page