himel7
/

bias-detector

@@ -1,199 +1,128 @@
 ---
 library_name: transformers
-tags: []
 ---
 # Model Card for Model ID
-<!-- Provide a quick summary of what the model is/does. -->
 ## Model Details
 ### Model Description
-<!-- Provide a longer summary of what this model is. -->
-This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
-- **Developed by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Model type:** [More Information Needed]
-- **Language(s) (NLP):** [More Information Needed]
-- **License:** [More Information Needed]
-- **Finetuned from model [optional]:** [More Information Needed]
-### Model Sources [optional]
-<!-- Provide the basic links for the model. -->
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
 ## Uses
-<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
-### Direct Use
-<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
-[More Information Needed]
-### Downstream Use [optional]
-<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
-[More Information Needed]
-### Out-of-Scope Use
-<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
-[More Information Needed]
-## Bias, Risks, and Limitations
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
-[More Information Needed]
-### Recommendations
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
-Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
-## How to Get Started with the Model
-Use the code below to get started with the model.
-[More Information Needed]
-## Training Details
-### Training Data
-<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
-[More Information Needed]
-### Training Procedure
-<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
-#### Preprocessing [optional]
-[More Information Needed]
-#### Training Hyperparameters
-- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
-#### Speeds, Sizes, Times [optional]
-<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
-[More Information Needed]
 ## Evaluation
-<!-- This section describes the evaluation protocols and provides the results. -->
-### Testing Data, Factors & Metrics
-#### Testing Data
-<!-- This should link to a Dataset Card if possible. -->
-[More Information Needed]
-#### Factors
-<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
-[More Information Needed]
-#### Metrics
-<!-- These are the evaluation metrics being used, ideally with a description of why. -->
-[More Information Needed]
-### Results
-[More Information Needed]
 #### Summary
-## Model Examination [optional]
-<!-- Relevant interpretability work for the model goes here -->
-[More Information Needed]
-## Environmental Impact
-<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
-Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
-- **Hardware Type:** [More Information Needed]
-- **Hours used:** [More Information Needed]
-- **Cloud Provider:** [More Information Needed]
-- **Compute Region:** [More Information Needed]
-- **Carbon Emitted:** [More Information Needed]
-## Technical Specifications [optional]
-### Model Architecture and Objective
-[More Information Needed]
-### Compute Infrastructure
-[More Information Needed]
-#### Hardware
-[More Information Needed]
-#### Software
-[More Information Needed]
-## Citation [optional]
-<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
-**BibTeX:**
-[More Information Needed]
-**APA:**
-[More Information Needed]
-## Glossary [optional]
-<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
-[More Information Needed]
-## More Information [optional]
-[More Information Needed]
-## Model Card Authors [optional]
-[More Information Needed]
-## Model Card Contact
-[More Information Needed]

 ---
 library_name: transformers
+tags:
+- Bias Detection
+- Text Classification
+language:
+- en
+Author:
+- Himel Ghosh
 ---
 # Model Card for Model ID
+This is a RoBERTa-based binary classification model fine-tuned on the BABE (URL: https://huggingface.co/datasets/mediabiasgroup/BABE)
+dataset for bias detection in English news statements.
+The model predicts whether a given sentence contains biased language (LABEL_1) or is unbiased (LABEL_0).
+It is intended for applications in media bias analysis, content moderation, and social computing research.
+## Citation
+Please cite this Hugging Face model page:
+@misc{himel7robertaBabe2025,
+title={RoBERTa Bias Detector (fine-tuned on BABE)},
+author={Himel Ghosh},
+howpublished={\url{https://huggingface.co/himel7/roberta-babe}},
+year={2025}
+}
 ## Model Details
 ### Model Description
+This model is a fine-tuned version of roberta-base trained to detect linguistic bias in English-language news statements.
+The task is framed as binary classification: the model outputs LABEL_1 for biased statements and LABEL_0 for non-biased statements.
+Fine-tuning was performed on the BABE dataset, which contains annotated news snippets across various topics and political leanings.
+The annotations focus on whether the language used expresses subjective bias rather than factual reporting.
+The goal of this model is to assist in detecting subtle forms of bias in media content, such as emotionally loaded language, stereotypical
+phrasing, or exaggerated claims, and can be useful in journalistic analysis, media monitoring, or NLP research into framing and stance.
+This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
+- **Developed by:** Himel Ghosh
+- **Language(s) (NLP):** Python
+- **Finetuned from model:** roberta-base
 ## Uses
+This model is intended to support the detection and analysis of biased language in English news content. It can be used as a tool by:
+- **Media researchers** and **social scientists** studying framing, bias, or political discourse.
+- **Journalists and editors** aiming to assess the neutrality of their writing or compare outlets.
+- **Developers** integrating bias detection into NLP pipelines for content moderation, misinformation detection, or AI-assisted writing tools.
+  ### Foreseeable Uses:
+- Annotating datasets for bias.
+- Measuring bias across different news outlets or topics.
+- Serving as an assistive tool in editorial decision-making or media monitoring.
+### Users Affected:
+- **Content creators** whose work may be labeled as biased or unbiased.
+- **End-users** of applications powered by this model (e.g., fact-checking platforms, moderation systems).
+- **Marginalized communities**, depending on how bias is defined, interpreted, and acted upon.
+The model should be used with care in high-stakes contexts, as bias is inherently subjective and culturally contextual.
+It is not intended to replace human judgment but to assist in surfacing potentially biased expressions.
+### Direct Use
+This model can be used directly for binary classification of English-language news statements to determine whether they exhibit biased language.
+It returns one of two labels:
+- **LABEL_0** –> Non-biased
+- **LABEL_1** –> Biased
+- Example usage with Hugging Face’s pipeline:
+  from transformers import pipeline
+  classifier = pipeline("text-classification", model="your-username/roberta-babe")
+  result = classifier("Immigrants are criminals.")
+## Bias, Risks, and Limitations
+While this model is designed to detect linguistic bias, it carries several limitations and risks, both technical and sociotechnical:
+- The model was fine-tuned on the BABE dataset, which includes annotations based on human judgments that may reflect specific cultural or political perspectives.
+- It may not generalize well to non-news text or out-of-domain content (e.g., social media, informal writing).
+- Subtle forms of bias, sarcasm, irony, or coded language may not be reliably detected.
+- Bias is inherently subjective: What one annotator considers biased may be seen as neutral by another. The model reflects those subjective judgments.
+- The model does not detect factual correctness or misinformation — only linguistic bias cues.
+- Labeling a text as “biased” may have reputational or ethical implications, especially if used in moderation, censorship, or journalistic evaluations.
+### Recommendations
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+## Training Details
+### Training Data
+Training was done on the BABE Dataset: https://huggingface.co/datasets/mediabiasgroup/BABE
 ## Evaluation
+The model was evaluated on the Test split of the BABE dataset and yielded the following metrics:
+- **Accuracy:  0.8520**
+- **Precision: 0.9237**
+- **Recall:    0.8014**
+- **F1 Score:  0.8582**
 #### Summary
+The model achieved 85.2% Accuracy on the BABE test split, with very high Precision of 92.37% and 80.14% Recall.
+This means the model predicts very few false positives and detects the biases that are actually biases.