WEKA ARFF文件预处理：为NLP情感分析语料加单引号及工具咨询

Hey there! Let's tackle this problem for your NLP sentiment analysis project. I've got two solid approaches for you: one with Python code to automate the task, and another set of no-code tools that get the job done quickly.

1. Python代码实现 #

ARFF files have metadata lines (starting with @) that we need to preserve, so the code will skip those and only modify actual data rows. Here's a straightforward script:

# Replace these paths with your actual file paths
input_arff = "your_input_data.arff"
output_arff = "modified_sentences.arff"

# Read and process the file
with open(input_arff, 'r', encoding='utf-8') as infile, open(output_arff, 'w', encoding='utf-8') as outfile:
    for line in infile:
        cleaned_line = line.strip()
        # Skip empty lines and ARFF metadata (like @relation, @attribute)
        if not cleaned_line or cleaned_line.startswith('@'):
            outfile.write(line)
            continue
        # Split text and label only once (avoids breaking sentences with commas)
        if ',1' in cleaned_line:
            text_segment, label_segment = cleaned_line.split(',', 1)
            # Wrap text in single quotes and reconstruct the line
            modified_line = f"'{text_segment}',{label_segment}\n"
            outfile.write(modified_line)
        else:
            # Keep lines with other labels (like 0) unchanged, or add similar logic if needed
            outfile.write(line)

Notes on the code: #

• Uses split(',', 1) to handle sentences that contain commas (no accidental splitting of the actual text)
• Preserves all original ARFF metadata so the file remains compatible with WEKA
• You can extend the logic to handle other labels (e.g., ,0) by copying the if ',1' block and adjusting the condition

2. No-Code Alternative Tools #

If you don't want to write code, these tools work perfectly:

Option 1: Notepad++ (Regex Replace) #

This is the fastest method for most cases:

1. Open your ARFF file in Notepad++
2. Press Ctrl+H to open the Replace window
3. Enable Regular Expression mode (check the box in the bottom-left corner)
4. In the "Find what" field, enter: ^([^,]+)(,1)$
5. In the "Replace with" field, enter: '\1'\2
6. Click Replace All

If you need to handle other labels (like ,0), adjust the regex to ^([^,]+)(,[01])$ — this will wrap text for both label 0 and 1.

Option 2: Excel (For Well-Formatted Data) #

Only use this if your sentences don't contain commas (otherwise Excel will split columns incorrectly):

1. Open Excel, go to Data > From Text/CSV and select your ARFF file
2. Choose comma as the delimiter, then skip all metadata rows (the ones starting with @)
3. In a new column, use the formula: =CONCATENATE("'",A1,"'",",",B1) (assuming text is in column A, label in B)
4. Copy the processed column, paste it into a new text file, add back the original ARFF metadata, and save as a new .arff file

内容的提问来源于stack exchange，提问作者A. Atiyah