tag-extractor

Python Flickr and Instagram tag extractor (by location) using Python 3

How to Run

Simply:

python3 setup.py develop

After setup please insteall To install the nltk modules, run the following python3 script:

#! /usr/bin/python3
import nltk
nltk.download("omw")
nltk.download("brown")

Then:

tagextractor --config YOURCONFIG.yml

Configuration

Example of configuration

--- # Configuration
extraction:
  enabled: true
  api: instagram
  api_cfg:
    ACCESS_TOKEN : YOUR_TOKEN
    ACCESS_SECRET : YOUR_SECRET_KEY
    CONSUMER_KEY : CONSUMER_KEY
    CONSUMER_SECRET : CONSUMER_SECRET
    frequency: ../resources/frequs.json
  location:
    lat: 46.205850
    lon: 6.157521
    radius: 1
  number: 1000
  pipeline:
    Babel: false
    WordNet: true
  storage:
    module: DB
    module_cfg:
      path: sqlite:///database/output.db
classification:
  enabled: true
  inputdb: sqlite:///database/output.db
  ontology_path: resources
  ontology: kr-owlxml.owl
  ontology_name: http://tagis.kr.com
  outputdb: sqlite:///database/output-classified.db

Architecture

Extraction

Classification

References and Links

Dependencies

The project uses multiple standard python libraries:

• nltk (and the modules omw and brown)
• unidecode
• SQLAlchemy
• flickrapi
• twitter
• requests
• PyYAML
• Owlready

References

• A methodology for mapping Instagram hashtags

Links

• PyBabelfy
• python twitter examples
• Piplines in Python
• FlickrAPI bug fix