data_analysis.py 文件源码

def get_relevant_entities(self, google_cloud_entities, target_entities, target_wikipedia_urls):
        entities_to_return = []
        target_wikipedia_urls_lower = [target_wikipedia_url.lower() for target_wikipedia_url in target_wikipedia_urls]

        for google_cloud_entity in google_cloud_entities:
            # Look at Wikipedia URLs
            if google_cloud_entity.wikipedia_url and google_cloud_entity.wikipedia_url.lower() in target_wikipedia_urls_lower:
                entities_to_return.append(google_cloud_entity.name)
                continue

            # Look at names
            a = google_cloud_entity.name.lower().split(" ")
            for target_entity in target_entities:
                b = target_entity.lower().split(" ")

                if google_cloud_entity in entities_to_return:
                    break

                for google_cloud_entity_part in a:
                    for target_entity_part in b:
                        ratio = SequenceMatcher(None, google_cloud_entity_part, target_entity_part).ratio()

                        if ratio > 0.7:
                            entities_to_return.append(google_cloud_entity.name)
                            break

                    if google_cloud_entity in entities_to_return:
                        break

        return entities_to_return