是否存在与R语言syuzhe包功能类似的Python情感分析库？

Great question! Since you're targeting a rule-based, lexicon-driven approach (no machine learning, perfect for unlabeled data) to detect the 8 NRC emotions—anger, anticipation, disgust, fear, joy, sadness, surprise, trust—here are your best Python options:

1. nrclex – The Direct Drop-In Equivalent #

This library is built specifically around the NRC Word-Emotion Association Lexicon, mirroring the core functionality of syuzhet's get_nrc_sentiment() function. It counts occurrences of each of the 8 target emotions in your text, no training required.

How to use it: #

First install the package:

pip install nrclex

Then implement it in your code:

from nrclex import NRCLex

# Example comment text
comment = "I'm stoked for the new game launch but terrified the servers will crash on day one!"

# Run emotion analysis
emotion_results = NRCLex(comment)

# Extract emotion counts (matches the 8 NRC categories + valence)
emotion_counts = emotion_results.affect_frequencies

# Print readable results
for emotion, count in emotion_counts.items():
    print(f"{emotion}: {count}")

This will output a breakdown of each emotion's presence, just like the R tool you're familiar with.

2. nltk + Custom NRC Lexicon Implementation #

If you prefer using a more established library like NLTK, you can manually load the NRC lexicon and build a custom function to calculate emotion counts. This gives you full control over preprocessing steps like tokenization or stopword removal.

Steps & Example Code: #

1. Download the public NRC Emotion Lexicon (available as a CSV/TXT file)
2. Load it into a Python dictionary mapping words to their associated emotions
3. Preprocess your text and count emotion matches

import nltk
from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
import pandas as pd

# Download required NLTK resources
nltk.download('punkt')
nltk.download('stopwords')

# Load NRC lexicon (adjust file path to your local copy)
nrc_df = pd.read_csv("nrc_lexicon.csv", names=["word", "emotion", "association"])
emotion_map = {}
for _, row in nrc_df.iterrows():
    if row["association"] == 1:
        if row["word"] not in emotion_map:
            emotion_map[row["word"]] = []
        emotion_map[row["word"]].append(row["emotion"])

# Text preprocessing helper
def clean_text(text):
    tokens = word_tokenize(text.lower())
    stop_words = set(stopwords.words('english'))
    return [token for token in tokens if token.isalpha() and token not in stop_words]

# Custom emotion counting function
def get_nrc_emotions(text):
    tokens = clean_text(text)
    emotion_counts = {
        "anger": 0, "anticipation": 0, "disgust": 0, "fear": 0,
        "joy": 0, "sadness": 0, "surprise": 0, "trust": 0
    }
    for token in tokens:
        if token in emotion_map:
            for emotion in emotion_map[token]:
                if emotion in emotion_counts:
                    emotion_counts[emotion] += 1
    return emotion_counts

# Test with sample text
comment = "This product is amazing! I trust the brand completely, though I was surprised by how fast it shipped."
print(get_nrc_emotions(comment))

3. text2emotion (Partial Alternative) #

While not strictly tied to the NRC lexicon, text2emotion is a simple out-of-the-box tool that detects similar core emotions. Note: it doesn't include anticipation, disgust, or trust, so it's only a partial fit for your full 8-emotion requirement.

Quick Usage: #

pip install text2emotion

import text2emotion as te

comment = "The customer service was terrible— I'm so angry they ignored my complaint for weeks!"
print(te.get_emotion(comment))

内容的提问来源于stack exchange，提问作者Christian Gonzalez-Martel