# Academic Query Classification — Data

来源：https://www.kaggle.com/competitions/academic-query-classification/data

采集时间：2026-10-20T04:30:05Z

## Files

*   **train.csv** - the training set
*   **test.csv** - the test set
*   **sample_submission.csv** - a sample submission file in the correct format

## Columns
unique ID	query_text	category
*   `unique ID` - Unique ID corresponding to the query text
*   `query_text` - query text
*  `category` - category of the query

## Details of the classification classes (categories):
**academic_servable:** Queries that indicate a clear intent to seek or provide academic services, such as tutoring, research assistance, or course-related help. e.g. 'what is inertia'.
**academic_non_servable:** Queries related to academic topics but without a clear intent to seek or provide services. These might include general academic discussions or queries about educational concepts. e.g. 'What will come in place of the question mark(?) in the given number series?'.
**general:** Non-academic queries that cover a wide range of topics unrelated to academics. e.g. 'Could you give me a brief overview of haryana police si'.
**junk:** Queries that contain irrelevant, nonsensical, or spam content. e.g. 'X-Men: Mutant Academy 2'.
**conversational:** Queries that are part of a conversation or dialogue and don't necessarily fall into the other categories. e.g. 'ansar me'.

**Note**
1. Training Dataset contains a total of 5k query texts and their categories tagged. 
2. Training data has an imbalance.

## Save final submission
Suppose your final data frame is `output`. Save the final output to the file 'submission.csv' containing only two columns 'unique ID' and 'category'.
`output[['unique ID', 'category']].to_csv('submission.csv', index=False)`
Then after saving the notebook version submit the output for evaluation.