Question Complexity Rating Dataset

Dataset Description

This dataset contains pairs of questions and their respective complexity ratings. The complexity ratings are provided on a scale from 0.0 to 1.0, where 0.0 indicates the simplest questions and 1.0 indicates the most complex questions. The dataset is intended to be used for training and evaluating models that classify or rank questions based on their complexity.

Dataset Contents

The dataset is provided as a CSV file with the following columns:

instruction: The question or prompt.
label: The complexity rating of the question, a float between 0.0 and 1.0.

Example

instruction	label
When did Virgin Australia start operating?	0.2
Which is a species of fish? Tope or Rope	0.8
Why can camels survive for long without water?	0.5

Usage

The dataset can be used for various natural language processing (NLP) tasks such as:

Complexity prediction
Question ranking
Model evaluation
Transfer learning for related tasks

Use Cases

1. Complexity Prediction

Train a model to predict the complexity of new questions based on the provided ratings. This can be useful in educational technology, customer support, and other domains where question complexity is a relevant factor.

2. Question Ranking

Rank a set of questions by their complexity to prioritize easier questions for beginners and more complex questions for advanced users.

3. Model Evaluation

Use the dataset to evaluate the performance of models designed to assess or rank question complexity.

How to Use the Dataset

To use the dataset, load the CSV file using a library like pandas and preprocess it as needed for your specific use case.

Loading the Dataset

import pandas as pd

# Load the dataset
df = pd.read_csv('path_to_your_dataset.csv')

# Display the first few rows of the dataset
print(df.head())

Training a Model

Here is a simple example of how to train a model using the dataset:

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

# Load the dataset
df = pd.read_csv('path_to_your_dataset.csv')

# Split the dataset into training and test sets
train_df, test_df = train_test_split(df, test_size=0.2, random_state=42)

# Extract features and labels
X_train = train_df['instruction']
y_train = train_df['label']
X_test = test_df['instruction']
y_test = test_df['label']

# Convert text data to numerical features (example using TF-IDF)
from sklearn.feature_extraction.text import TfidfVectorizer
vectorizer = TfidfVectorizer()

X_train_tfidf = vectorizer.fit_transform(X_train)
X_test_tfidf = vectorizer.transform(X_test)

# Train a simple model (example using Linear Regression)
model = LinearRegression()
model.fit(X_train_tfidf, y_train)

# Evaluate the model
predictions = model.predict(X_test_tfidf)
print(f"Predictions: {predictions}")

Citation If you use this dataset in your research, please cite it as follows:

@dataset{question_complexity_rating_2024,
  title = {Question Complexity Rating Dataset},
  author = {Wes Lagarde},
  year = {2024},
  url = {https://huggingface.co/wesley7137/question_complexity_classification},
  description = {A dataset of question and complexity rating pairs for natural language processing tasks.}
}

License This dataset is licensed under the Apache 2.0 License.

Acknowledgements We would like to thank all contributors and the community for their valuable feedback and support.

css Copy code

This README file provides a comprehensive overview of your dataset, including its contents, usage, and example code for loading and using the dataset. Adjust the content as necessary to fit your specific dataset details and preferences.