能否输出sentiment score

[TaoRuan-Campus]

你好，请问预测英文句子级情感分类任务当中，除了Positive和negative，是否可以同时输出sentiment score进行比较？

[LucienShui]

import numpy as np
from senta import Senta
from senta.common.rule import InstanceName
from senta.data.util_helper import convert_texts_to_ids, structure_fields_dict
from senta.utils.util_helper import array2tensor, text_type


class SentimentChineseClassifier(Senta):
    def __init__(self):
        super(SentimentChineseClassifier, self).__init__()
        self.init_model()

    def predict(self, texts_, aspects=None):
        if isinstance(texts_, text_type):
            texts_ = [texts_]

        if isinstance(aspects, text_type):
            aspects = [aspects]

        return_list = convert_texts_to_ids(
            texts_, self.tokenizer, self.max_seq_len, self.truncation_type, self.padding_id)
        record_dict = structure_fields_dict(return_list, 0, need_emb=False)
        input_list = []
        for item in self.input_keys:
            kv = item.split("#")
            name = kv[0]
            key = kv[1]
            input_item = record_dict[InstanceName.RECORD_ID][key]
            input_list.append(input_item)
        inputs = [array2tensor(ndarray) for ndarray in input_list]
        result = self.inference.run(inputs)
        batch_result = self.model_class.parse_predict_result(result)
        results = []
        if self.inference_type == 'seq_lab':
            for text, probs in zip(texts_, batch_result):
                label = [self.label_map[l] for l in probs]
                results.append((text, label, probs.tolist()))
        else:
            for text, probs in zip(texts_, batch_result):
                label = self.label_map[np.argmax(probs)]
                results.append((text, label, probs.tolist()))
        return results


def main():
    classifier = SentimentChineseClassifier()
    texts = ["PHP 是世界上最好的语言"]
    result = classifier.predict(texts)
    print(result)


if __name__ == '__main__':
    main()

可以参考下这段 Sample Code。

[TaoRuan-Campus]

import numpy as np
from senta import Senta
from senta.common.rule import InstanceName
from senta.data.util_helper import convert_texts_to_ids, structure_fields_dict
from senta.utils.util_helper import array2tensor, text_type


class SentimentChineseClassifier(Senta):
    def __init__(self):
        super(SentimentChineseClassifier, self).__init__()
        self.init_model()

    def predict(self, texts_, aspects=None):
        if isinstance(texts_, text_type):
            texts_ = [texts_]

        if isinstance(aspects, text_type):
            aspects = [aspects]

        return_list = convert_texts_to_ids(
            texts_, self.tokenizer, self.max_seq_len, self.truncation_type, self.padding_id)
        record_dict = structure_fields_dict(return_list, 0, need_emb=False)
        input_list = []
        for item in self.input_keys:
            kv = item.split("#")
            name = kv[0]
            key = kv[1]
            input_item = record_dict[InstanceName.RECORD_ID][key]
            input_list.append(input_item)
        inputs = [array2tensor(ndarray) for ndarray in input_list]
        result = self.inference.run(inputs)
        batch_result = self.model_class.parse_predict_result(result)
        results = []
        if self.inference_type == 'seq_lab':
            for text, probs in zip(texts_, batch_result):
                label = [self.label_map[l] for l in probs]
                results.append((text, label, probs.tolist()))
        else:
            for text, probs in zip(texts_, batch_result):
                label = self.label_map[np.argmax(probs)]
                results.append((text, label, probs.tolist()))
        return results


def main():
    classifier = SentimentChineseClassifier()
    texts = ["PHP 是世界上最好的语言"]
    result = classifier.predict(texts)
    print(result)


if __name__ == '__main__':
    main()

可以参考下这段 Sample Code。

非常感谢！我会仔细研究这个案例。另外我想问一下，如果是对英文文本进行处理，应该在哪里改动以使用对应的英文预训练模型呢？

[LucienShui]

改动 self.init_model() 这一行，使用英文模型初始化即可。