def suggest_locations():
"""
Suggest locations in a text string. These might be useful keywords for
annotators to geolocate.
input: full message's text [string]
output: dict of with outer key "locations". inner keys are entity types, with list of entities.
"locations": {
"GSP": [
"Congo"
]
}
# [TODO]
# output: list. each item is a python dictionary:
# - text : the text for the specific entity [string]
# - indices : tuple of (start [int], end [int]) offset where entity is
# located in given full message
# - confidence : probability from 0-to-1 [float]
"""
if not request.json and not 'text' in request.json:
abort(400)
# Get all entities and only fetch GPE
entities = Machine.guess_locations(request.json['text'])
for k, v in entities.iteritems():
entities[k] = list(v)
return jsonify({'locations': entities})
def suggest_locations():
"""
Suggest locations in a text string. These might be useful keywords for
annotators to geolocate.
input: full message's text [string]
output: dict of with outer key "locations". inner keys are entity types, with list of entities.
"locations": {
"GSP": [
"Congo"
]
}
# [TODO]
# output: list. each item is a python dictionary:
# - text : the text for the specific entity [string]
# - indices : tuple of (start [int], end [int]) offset where entity is
# located in given full message
# - confidence : probability from 0-to-1 [float]
"""
if not request.json and not 'text' in request.json:
abort(400)
# Get all entities and only fetch GPE
entities = Machine.guess_locations(request.json['text'])
for k, v in entities.iteritems():
entities[k] = list(v)
return jsonify({'locations': entities})
def detect_language():
"""Given some text, returns a ranked list of likey natural languages
the given content is in
Input parameters:
text: string
"""
if not request.json or not 'text' in request.json:
abort(400)
language = Machine.guess_language(request.json['text'])[0]
return jsonify({'language': language[0], "confidence": language[1]})
def detect_language():
"""Given some text, returns a ranked list of likey natural languages
the given content is in
Input parameters:
text: string
"""
if not request.json or not 'text' in request.json:
abort(400)
language = Machine.guess_language(request.json['text'])[0]
return jsonify({'language': language[0], "confidence": language[1]})
def extract_entities():
"""Given some text input, identify - besides location - people,
organisations and other types of entities within the text"""
pass
if not request.json and not 'text' in request.json:
abort(400)
result = Machine.guess_entities(request.json['text'])
entities = {}
for key, value in result.iteritems():
entities[key.lower()] = list(value)
return jsonify({'entities': entities})
def extract_entities():
"""Given some text input, identify - besides location - people,
organisations and other types of entities within the text"""
pass
if not request.json and not 'text' in request.json:
abort(400)
result = Machine.guess_entities(request.json['text'])
entities = {}
for key, value in result.iteritems():
entities[key.lower()] = list(value)
return jsonify({'entities': entities})
def suggest_sensitive_info():
"""
Suggest personally identifying information (PII) -- such as
credit card numbers, phone numbers, email, etc --
from a text string. These are useful for annotators to investigate
and strip before publicly posting information.
input: text,
input: options
- custom regex for local phone numbers
- flags or booleans to specify the type of pii (e.g. phone_only)
output: list of dictionaries:
- word
- type (e-mail, phone, ID, person name, etc.)
- indices (start/end offset in text)
- confidence [todo: is possible?]
"""
if not request.json and not 'text' in request.json:
abort(400)
private_info = Machine.guess_private_info(request.json['text'])
return jsonify({'private_info': private_info})
def suggest_sensitive_info():
"""
Suggest personally identifying information (PII) -- such as
credit card numbers, phone numbers, email, etc --
from a text string. These are useful for annotators to investigate
and strip before publicly posting information.
input: text,
input: options
- custom regex for local phone numbers
- flags or booleans to specify the type of pii (e.g. phone_only)
output: list of dictionaries:
- word
- type (e-mail, phone, ID, person name, etc.)
- indices (start/end offset in text)
- confidence [todo: is possible?]
"""
if not request.json and not 'text' in request.json:
abort(400)
private_info = Machine.guess_private_info(request.json['text'])
return jsonify({'private_info': private_info})