relation_extraction.py

import evaluate
import datasets
import numpy as np

# Add BibTeX citation
_CITATION = """\
@Paper{
    author = {Bruno Taillé, Vincent Guigue, Geoffrey Scoutheeten, Patrick Gallinari},
    title = {Let's Stop Incorrect Comparisons in End-to-end Relation Extraction!},
    year = {2020},
    link = https://arxiv.org/abs/2009.10684
}
"""

# Add description of the module here
_DESCRIPTION = """\
This metric is used for evaluating the quality of relation extraction output. By calculating the Micro and Macro F1 score of every relation extraction outputs to ensure the quality.
"""


# Add description of the arguments of the module here
_KWARGS_DESCRIPTION = """
Calculates how good are predictions given some references, using Micro and Macro F1 scores
Args:
    predictions: list of list of dictionary, including relation and its type 
        dictionary should be key value pair like entity name link to its type
    references: list of list of dictionary, including relation and its type 
        dictionary should be entity name pair like entity name link to its type
Returns:
    evaluation result: 
    - **sell** (`dictionary`): score of type sell
  - **tp** : true positive count
  - **fp** : false positive count
  - **fn** : false negative count
  - **p** : precision
  - **r** : recall
  - **f1** : micro f1 score
- **ALL** (`dictionary`): score of all of the type (sell and belongs to)
  - **tp** : true positive count
  - **fp** : false positive count
  - **fn** : false negative count
  - **p** : precision
  - **r** : recall
  - **f1** : micro f1 score
  - **Macro_f1** : macro f1 score
  - **Macro_p** : macro precision
  - **Macro_r** : macro recall
Examples:
    >>> metric_path = "Ikala-allen/relation_extraction"
    >>> module = evaluate.load(metric_path)
    >>> references = [
    ...    [
    ...       {"head": "phip igments", "head_type": "brand", "type": "sell", "tail": "國際認證之色乳", "tail_type": "product"},
    ...        {"head": "tinadaviespigments", "head_type": "brand", "type": "sell", "tail": "國際認證之色乳", "tail_type": "product"},
    ...    ]
    ... ]
    >>> predictions = [
    ...    [
    ...        {"head": "phipigments", "head_type": "product", "type": "sell", "tail": "國際認證之色乳", "tail_type": "product"},
    ...        {"head": "tinadaviespigments", "head_type": "brand", "type": "sell", "tail": "國際認證之色乳", "tail_type": "product"},
    ...    ]
    ... ]
    >>> evaluation_scores = module.compute(predictions=predictions, references=references)
    >>> print(evaluation_scores)
    {'sell': {'tp': 1, 'fp': 1, 'fn': 1, 'p': 50.0, 'r': 50.0, 'f1': 50.0}, 'ALL': {'tp': 1, 'fp': 1, 'fn': 1, 'p': 50.0, 'r': 50.0, 'f1': 50.0, 'Macro_f1': 50.0, 'Macro_p': 50.0, 'Macro_r': 50.0}}
"""

def convert_format(data:list):
    """
    Args:
        data (list) : list of dictionaries with different entity elements
    e.g
        [
        {'head': ['phipigments', 'tinadaviespigments'...], 
        'head_type': ['product', 'brand'...], 
        'type': ['sell', 'sell'...], 
        'tail': ['國際認證之色乳', '國際認證之色乳'...], 
        'tail_type': ['product', 'product'...]},  # first element
        {'head': ['SABONTAIWAN', 'SNTAIWAN'...], 
        'head_type': ['brand', 'brand'...], 
        'type': ['sell', 'sell'...], 
        'tail': ['大馬士革玫瑰有機光燦系列', '大馬士革玫瑰有機光燦系列'...], 
        'tail_type': ['product', 'product'...]}  # second element
        ...
        ]
    """
    predictions = []
    for item in data:
        prediction_group = []
        for i in range(len(item['head'])):
            prediction = {
                'head': item['head'][i],
                'head_type': item['head_type'][i],
                'type': item['type'][i],
                'tail': item['tail'][i],
                'tail_type': item['tail_type'][i],
            }
            prediction_group.append(prediction)
        predictions.append(prediction_group)
    return predictions

@evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class relation_extraction(evaluate.Metric):
    """
    evaluation metric of relation extraction
    inputs:
        predictions : (`list` of `list`s of `dictionary`s) about relation and its type of prediction
        references : (`list` of `list`s of `dictionary`s) about references for each relation and its type.
    """

    def _info(self):
        # TODO: Specifies the evaluate.EvaluationModuleInfo object
        return evaluate.MetricInfo(
            # This is the description that will appear on the modules page.
            module_type="metric",
            description=_DESCRIPTION,
            citation=_CITATION,
            inputs_description=_KWARGS_DESCRIPTION,
            # This defines the format of each prediction and reference
            features=datasets.Features({
                'predictions': datasets.Sequence({
                    "head": datasets.Value("string"),
                    "head_type": datasets.Value("string"),
                    "type": datasets.Value("string"),
                    "tail": datasets.Value("string"),
                    "tail_type": datasets.Value("string"),
                }),
                'references': datasets.Sequence({
                    "head": datasets.Value("string"),
                    "head_type": datasets.Value("string"),
                    "type": datasets.Value("string"),
                    "tail": datasets.Value("string"),
                    "tail_type": datasets.Value("string"),
                }),
            }),
            # Homepage of the module for documentation
            homepage="http://module.homepage",
            # Additional links to the codebase or references
            codebase_urls=["http://github.com/path/to/codebase/of/new_module"],
            reference_urls=["http://path.to.reference.url/new_module"]
        )

    def _download_and_prepare(self, dl_manager):
        """Optional: download external resources useful to compute the scores"""
        # TODO: Download external resources if needed
        pass

    def _compute(self, predictions, references, mode="strict", relation_types=[]):
        """Returns the scores"""
        # TODO: Compute the different scores of the module
        
        predictions = convert_format(predictions)
        references = convert_format(references)
        
        assert mode in ["strict", "boundaries"]

        # construct relation_types from ground truth if not given
        if len(relation_types) == 0:
            for triplets in references:
                for triplet in triplets:
                    relation = triplet["type"]
                    if relation not in relation_types:
                        relation_types.append(relation)
        
        scores = {rel: {"tp": 0, "fp": 0, "fn": 0} for rel in relation_types + ["ALL"]}
        
        # Count GT relations and Predicted relations
        n_sents = len(references)
        n_rels = sum([len([rel for rel in sent]) for sent in references])
        n_found = sum([len([rel for rel in sent]) for sent in predictions])

        # Count TP, FP and FN per type
        for pred_sent, gt_sent in zip(predictions, references):
            for rel_type in relation_types:
                # strict mode takes argument types into account
                if mode == "strict":
                    pred_rels = {(rel["head"], rel["head_type"], rel["tail"], rel["tail_type"]) for rel in pred_sent if
                                rel["type"] == rel_type}
                    gt_rels = {(rel["head"], rel["head_type"], rel["tail"], rel["tail_type"]) for rel in gt_sent if
                            rel["type"] == rel_type}

                # boundaries mode only takes argument spans into account
                elif mode == "boundaries":
                    pred_rels = {(rel["head"], rel["tail"]) for rel in pred_sent if rel["type"] == rel_type}
                    gt_rels = {(rel["head"], rel["tail"]) for rel in gt_sent if rel["type"] == rel_type}

                scores[rel_type]["tp"] += len(pred_rels & gt_rels)
                scores[rel_type]["fp"] += len(pred_rels - gt_rels)
                scores[rel_type]["fn"] += len(gt_rels - pred_rels)

        # Compute per entity Precision / Recall / F1
        for rel_type in scores.keys():
            if scores[rel_type]["tp"]:
                scores[rel_type]["p"] = 100 * scores[rel_type]["tp"] / (scores[rel_type]["fp"] + scores[rel_type]["tp"])
                scores[rel_type]["r"] = 100 * scores[rel_type]["tp"] / (scores[rel_type]["fn"] + scores[rel_type]["tp"])
            else:
                scores[rel_type]["p"], scores[rel_type]["r"] = 0, 0

            if not scores[rel_type]["p"] + scores[rel_type]["r"] == 0:
                scores[rel_type]["f1"] = 2 * scores[rel_type]["p"] * scores[rel_type]["r"] / (
                        scores[rel_type]["p"] + scores[rel_type]["r"])
            else:
                scores[rel_type]["f1"] = 0

        # Compute micro F1 Scores
        tp = sum([scores[rel_type]["tp"] for rel_type in relation_types])
        fp = sum([scores[rel_type]["fp"] for rel_type in relation_types])
        fn = sum([scores[rel_type]["fn"] for rel_type in relation_types])
        

        if tp:
            precision = 100 * tp / (tp + fp)
            recall = 100 * tp / (tp + fn)
            f1 = 2 * precision * recall / (precision + recall)

        else:
            precision, recall, f1 = 0, 0, 0

        scores["ALL"]["p"] = precision
        scores["ALL"]["r"] = recall
        scores["ALL"]["f1"] = f1
        scores["ALL"]["tp"] = tp
        scores["ALL"]["fp"] = fp
        scores["ALL"]["fn"] = fn


        # Compute Macro F1 Scores
        scores["ALL"]["Macro_f1"] = np.mean([scores[ent_type]["f1"] for ent_type in relation_types])
        scores["ALL"]["Macro_p"] = np.mean([scores[ent_type]["p"] for ent_type in relation_types])
        scores["ALL"]["Macro_r"] = np.mean([scores[ent_type]["r"] for ent_type in relation_types])

        return scores