From 5509e5c90f50c52bc42122cc5bcee66f4b7dade7 Mon Sep 17 00:00:00 2001
From: Ali Al Jasim <103039350+im4li@users.noreply.github.com>
Date: Wed, 1 May 2024 14:09:41 +0000
Subject: [PATCH 1/2] using Logistic Regression and similar to the example
 provided

---
 .../.dockerignore                             |  1 +
 authorship-verification-submission/Dockerfile |  5 ++
 .../authorship_verification_submission.py     | 62 +++++++++++++++++++
 3 files changed, 68 insertions(+)
 create mode 100644 authorship-verification-submission/.dockerignore
 create mode 100644 authorship-verification-submission/Dockerfile
 create mode 100644 authorship-verification-submission/authorship_verification_submission.py

diff --git a/authorship-verification-submission/.dockerignore b/authorship-verification-submission/.dockerignore
new file mode 100644
index 0000000..8c785a3
--- /dev/null
+++ b/authorship-verification-submission/.dockerignore
@@ -0,0 +1 @@
+predictions.jsonl
\ No newline at end of file
diff --git a/authorship-verification-submission/Dockerfile b/authorship-verification-submission/Dockerfile
new file mode 100644
index 0000000..1130b34
--- /dev/null
+++ b/authorship-verification-submission/Dockerfile
@@ -0,0 +1,5 @@
+FROM fschlatt/natural-language-processing-exercises:0.0.1
+
+ADD authorship_verification_submission.py /code/authorship_verification_submission.py
+
+ENTRYPOINT [ "python3", "/code/authorship_verification_submission.py" ]
\ No newline at end of file
diff --git a/authorship-verification-submission/authorship_verification_submission.py b/authorship-verification-submission/authorship_verification_submission.py
new file mode 100644
index 0000000..2c51851
--- /dev/null
+++ b/authorship-verification-submission/authorship_verification_submission.py
@@ -0,0 +1,62 @@
+from pathlib import Path
+
+from tira.rest_api_client import Client
+from tira.third_party_integrations import get_output_directory
+
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.linear_model import LogisticRegression
+from sklearn.metrics import accuracy_score
+from sklearn.pipeline import Pipeline
+
+
+
+
+def evaluate_model(model, data, labels):
+    predictions = model.predict(data)
+    accuracy = accuracy_score(labels, predictions)
+    return accuracy
+
+
+
+if __name__ == "__main__":
+    tira = Client()
+
+    # loading train data
+    text_train = tira.pd.inputs(
+        "nlpbuw-fsu-sose-24", "authorship-verification-train-20240408-training"
+    )
+    targets_train = tira.pd.truths(
+        "nlpbuw-fsu-sose-24", "authorship-verification-train-20240408-training"
+    )
+    # loading validation data (automatically replaced by test data when run on tira)
+    text_validation = tira.pd.inputs(
+        "nlpbuw-fsu-sose-24", "authorship-verification-validation-20240408-training"
+    )
+    targets_validation = tira.pd.truths(
+        "nlpbuw-fsu-sose-24", "authorship-verification-validation-20240408-training"
+    )
+
+    tfidf_vectorizer = TfidfVectorizer(max_features=1000)
+
+    # Model Training
+    model = Pipeline([
+        ('vectorizer', tfidf_vectorizer),
+        ('classifier', LogisticRegression())
+    ])
+
+    model.fit(text_train['text'], targets_train['generated'])
+
+
+    val_accuracy = evaluate_model(model, text_validation['text'], targets_validation['generated'])
+    # print("Validation Accuracy:", val_accuracy)
+    
+    # make predictions
+    predictions = model.predict(text_validation["text"])
+    text_validation["generated"] = predictions
+    df = text_validation[["id", "generated"]]
+
+    # Save the predictions
+    output_directory = get_output_directory(str(Path(__file__).parent))
+    df.to_json(
+        Path(output_directory) / "predictions.jsonl", orient="records", lines=True
+    )
\ No newline at end of file

From 063a124fa94c38cefe0016e038864902009d111c Mon Sep 17 00:00:00 2001
From: Ali Al Jasim <103039350+im4li@users.noreply.github.com>
Date: Wed, 1 May 2024 14:22:59 +0000
Subject: [PATCH 2/2] problem with output, trying to remove docker ignore

---
 authorship-verification-submission/.dockerignore | 1 -
 1 file changed, 1 deletion(-)

diff --git a/authorship-verification-submission/.dockerignore b/authorship-verification-submission/.dockerignore
index 8c785a3..e69de29 100644
--- a/authorship-verification-submission/.dockerignore
+++ b/authorship-verification-submission/.dockerignore
@@ -1 +0,0 @@
-predictions.jsonl
\ No newline at end of file