genderize.py

def genderize_api_connect(first_name):
    url = 'https://api.genderize.io/?name={}'.format(first_name)
    response = requests.get(url)
    try:
         assert response.status_code == 200
         content = json.loads(response.content)
    except AssertionError:
         self.logger.error("There was an error contacting the API service")
    return content

def author_gender(all_authors, threshold=0.6):
    for key, value in all_authors.items():
        # Ensure the author being evaluated has not been enriched using the API — this avoids unnecessary API calls
        if 'gender' not in value:
            try:
               # Retrieve the probability of the gender prediction of the author 
               # and compare it to the threshold set in line 11.
                result = genderize_api_connect(value['first_name'])
                if result['probability'] > threshold:
                    # Add the gender to the author dictionary object and ensure that it is capitalized.
                    value['gender'] = result['gender'].capitalize()
                    # Handle exceptions, specifically if no gender is returned from the API. 
                    # This case is handled by assigning the gender as “Unknown”.
            except KeyError:
                self.logger.error(
                    'Name of Author unknown for author_id {} in Research Paper with doi link {}'.format(key,
                                                                                                        value[
                                                                                                            'doi']))
                value['gender'] = "Unknown"
    return all_authors