nlp_architect.solutions.absa_solution package

Submodules

nlp_architect.solutions.absa_solution.sentiment_solution module

class nlp_architect.solutions.absa_solution.sentiment_solution.SentimentSolution(anonymiser: nlp_architect.solutions.absa_solution.utils.Anonymiser = None, max_events: int = 400)[source]

Bases: object

Main class for executing Sentiment Solution pipeline.

Parameters:
  • anonymiser (Anonymiser, optional) – Method to anonymise events’ text.
  • max_events (int, optional) – Maximum number of events to show for each aspect-polarity pair.
run(aspect_lex: os.PathLike = None, opinion_lex: os.PathLike = None, data: os.PathLike = None, parsed_data: os.PathLike = None, inference_results: os.PathLike = None) → Optional[pandas.core.frame.DataFrame][source]
nlp_architect.solutions.absa_solution.sentiment_solution.main() → None[source]

nlp_architect.solutions.absa_solution.ui module

nlp_architect.solutions.absa_solution.ui.empty_table(*headers)[source]
nlp_architect.solutions.absa_solution.ui.serve_absa_ui() → None[source]

Main function for serving UI application.

nlp_architect.solutions.absa_solution.utils module

class nlp_architect.solutions.absa_solution.utils.Anonymiser[source]

Bases: object

Abstract class for anonymiser algorithm, intended for privacy keeping.

run(text)[source]
class nlp_architect.solutions.absa_solution.utils.TweetAnonymiser(lexicon_path)[source]

Bases: nlp_architect.solutions.absa_solution.utils.Anonymiser

Anonymiser for tweets which uses lexicon for simple string replacements.

run(text)[source]

Module contents