Events¶
Introduction¶
The event detection module analyses a body of text to identify events that have taken place. This is useful for identifying mentions of personal experience or story mining.
Usage¶
To use the module:
>>> import coast_core
>>> coast_core.events.function(to_use)
or:
>>> from coast_core import events
>>> events.function(to_use)
Functions¶
Extract event instances from text.
-
coast_core.events.
get_iverb_bigrams
(text)¶ - Split a text into bigrams, following the pattern (“i”, <<verb>>).
- VB - Verb, base form VBD - Verb, past tense VBG - Verb, gerund or present participle VBN - Verb, past participle - n VBP - Verb, non-3rd person singular present - n VBZ - Verb, 3rd person singular present - n
Parameters: text – The text to analyse Returns: A dictionary containing ‘I verb’ bigrams and the total number of ‘I verb’ events.
-
coast_core.events.
get_timex_events
(text)¶ Given a body of text, returns a list of Timex events. Timex events are temporal events that are detected using regular expressions. Our timex library is a variation of the Timex module in NLTK_contrib: https://github.com/nltk/nltk_contrib/blob/master/nltk_contrib/timex.py
Parameters: text – The text to operate on Returns: return the timex events
-
coast_core.events.
get_verb_events
(text)¶ - Given a body of text, returns a list of verb events.
- VB - Verb, base form VBD - Verb, past tense VBG - Verb, gerund or present participle VBN - Verb, past participle - n VBP - Verb, non-3rd person singular present - n VBZ - Verb, 3rd person singular present - n
Parameters: text – The text to analyse. Returns: The list of verb events.
-
coast_core.events.
run_all_event_analysis
(article_text)¶ Run all event analysis for all articles.
Parameters: article_text – The text to analyse. Returns: An object containing timex events, verb events and iverb bigrams
-
coast_core.events.
timex_tag
(text, **kwargs)¶ Extract the timex events from a given body of text
Parameters: text – The body of text to operate on Returns: A list of timex events as default, unless ‘markup’ argument is given. In which case, returns a markedup string.