process4detailed.py

# /usr/bin/env python
# coding=utf-8
from collections import defaultdict
import json
import pandas as pd


def is_soo(triple):
    if (triple[0] in triple[-1]) or (triple[-1] in triple[0]):
        return True
    else:
        return False


def overlapping_search(triples):
    pattern = set()
    entity_pair = [(tri[0], tri[-1]) for tri in triples]
    if len(set(entity_pair)) != len(triples):
        pattern.add('EPO')

    entities = set()
    for idx, triple in enumerate(triples):
        if is_soo(triple):
            pattern.add('SOO')
        entities.add(triple[0])
        entities.add(triple[-1])
    if len(entities) != len(set(entity_pair)) * 2:
        pattern.add('SEO')

    if len(pattern) == 0:
        return ['Normal']
    else:
        return list(pattern)


def process():
    corpus_type = ['NYT-star', 'WebNLG-star', 'NYT', 'WebNLG']
    for t in corpus_type:
        with open(f'./{t}/test_triples.json', 'r', encoding='utf-8') as f:
            data = json.load(f)

        patterns = defaultdict(list)
        for idx, d in enumerate(data):
            pattern = overlapping_search(d['triple_list'])
            for pat in pattern:
                patterns[pat].append(idx)
        print(f'Patterns of {t}:', {k: len(v) for k, v in patterns.items()})
        for typ in patterns.keys():
            with open(f'./{t}/{typ}_triples.json', 'w', encoding='utf-8') as f:
                json.dump(pd.Series(data)[patterns[typ]].tolist(), f, indent=4, ensure_ascii=False)

        num_dict = defaultdict(list)
        for d in data:
            if len(d['triple_list']) >= 5:
                num_dict[5].append(d)
            else:
                num_dict[len(d['triple_list'])].append(d)
        for key, values in num_dict.items():
            with open(f'./{t}/{key}_triples.json', 'w', encoding='utf-8') as f:
                json.dump(values, f, indent=4, ensure_ascii=False)
        print(f'Triples of {t}:', {k: len(v) for k, v in num_dict.items()})


if __name__ == '__main__':
    process()