Update README.md

README.md

@@ -1,3 +1,73 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+---
+CISA-testset: Cross-Individual Sentiment Analysis Test Dataset for Historical Turkish
+This test dataset is specifically designed for evaluating Cross-Individual Sentiment Analysis (CISA) performance on historical Turkish texts from İbrahim Temo's memoirs.
+📚 Dataset Description
+This test dataset contains 200 sentences extracted from the first 66 pages of İbrahim Temo's original memoirs "İttihad ve Terakki Cemiyetinin Teşekkülü ve Hidematı Vataniye ve İnkılâbı Milliye Dair Hatıratım" published in 1939.
+Important Note: This is NOT part of the main training dataset used for the CISA-BERTurk-sentiment model. It serves as an independent test set to evaluate model performance on real historical texts.
+🎯 Purpose
+
+Independent Testing: Evaluate CISA model performance on unseen historical data
+Real-world Validation: Test model capabilities on authentic historical Turkish texts
+Benchmark Creation: Provide a standardized test set for CISA research
+
+📊 Dataset Statistics
+
+Total Sentences: 202
+Source: İbrahim Temo's Memoirs (1939) - First 66 pages
+Language: Historical Turkish (Early 20th century)
+Annotation Type: Cross-Individual Sentiment Analysis
+Entity Focus: Historical figures and personalities
+
+Sentiment Distribution
+
+Negative: 17 instances (8.4%)
+Neutral: 154 instances (76.2%)
+Positive: 31 instances (15.4%)
+
+Relation Type Distribution
+
+Direct: 174 instances (86.1%)
+Indirect: 28 instances (13.9%)
+
+🏆 Model Performance
+When tested with the CISA-BERTurk-sentiment model, this test set achieved superior performance compared to the original training test scores:
+Sentiment Analysis Results
+
+Accuracy: 88.12%
+Precision: 87.72%
+Recall: 88.12%
+F1-Score: 87.75%
+
+Key Finding: These results demonstrate that the model performs better on continuous, real-world historical texts compared to fragmented training data, indicating strong generalization capabilities.
+
+🏗️ Dataset Creation
+Dataset Prepared by: İbrahim Berci
+The dataset was carefully curated to:
+
+Maintain historical authenticity
+Ensure diverse entity-sentiment relationships
+Preserve original Ottoman Turkish linguistic features
+Provide comprehensive metadata for research reproducibility
+
+📚 Source Material
+Original Work: "İttihad ve Terakki Cemiyetinin Teşekkülü ve Hidematı Vataniye ve İnkılâbı Milliye Dair Hatıratım"
+
+Author: İbrahim Temo
+Publication Year: 1939
+Location: Mecidiye, Romania
+Content: Political memoirs of the Young Turk movement
+Language: Ottoman Turkish transitioning to modern Turkish
+
+🏦 Funding & Acknowledgments
+This work was supported by The Scientific and Technological Research Council of Turkey (TÜBİTAK) under project number 323K372. We thank TÜBİTAK for their support.
+
+Citation
+bibtex@dataset{berci2025cisa_testset,
+    author = {İbrahim Berci and Sezen Karabulut and Mustafa İlter},
+    title = {CISA-testset from Ibrahim Temo's memoir: Cross-Individual Sentiment Analysis Test Dataset for Historical Turkish},
+    url = {https://huggingface.co/datasets/dbbiyte/CISA-testset},
+    year = {2025},
+    note = {Test dataset for evaluating CISA performance on İbrahim Temo's historical memoirs}
+}