Add files via upload

.env

@@ -0,0 +1 @@
+OPENAI_API_KEY="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"

output.csv

@@ -0,0 +1,41 @@
+Tweet
+"@tedcruz And, #HandOverTheServer she wiped clean + 30k deleted emails, explains dereliction of duty/lies re #Benghazi,etc #tcot"
+Hillary is our best choice if we truly want to continue being a progressive nation. #Ohio
+"@TheView I think our country is ready for a female pres, it can't ever be Hillary"
+I just gave an unhealthy amount of my hard-earned money away to the big gov't & untrustworthy IRS. #WhyImNotVotingForHillary
+@PortiaABoulger Thank you for adding me to your list
+Hillary can not win. Here's hoping the Dems offer a real candidate like Warren. #Warren2016
+"Respect FOR the law and respect BY the law Yes, needed desperately.  #BaltimoreRiots"
+I don't want to be appointed to an Ambassador post.
+"#StopHillary2016 @HillaryClinton if there was a woman with integrity and honesty I would vote for such as woman president, NO"
+@HillaryClinton End lawless #ClintonFoundation. Jail Butcher of #Benghazi. #Arrest rapist #BillClinton. #HillaryClinton
+"Use your brain, keep Hillary out of the White House.Clinton2016"
+@HillaryClinton Hillary pandering with her logo. #ClintonFoundationscandal #ClintonCash
+"@readyforHRC @HillaryClinton #HillaryClinton, the US presidency is a testament to the success of #women their role in the world"
+@CiaraAntaya cuz you know I'm such a feminist
+2 million bogus followers on Twitter @HillaryClinton #WhyImNotVotingForHillary
+@lindasuhler : My name is Rebecca and my grandmother immigrated to Sunnybrook Farm. @twitchyteam
+Where's the campaign store is the real question? I am ready to buy some Hillary gear
+"It's a miracle, suddenly #Democrats don't mind having someone who voted for war."
+@smileitsalicia @greekgummybear2 now i can live in peace
+Hillary doesn't want to put anyone in prison anymore. Obviously worried about her own future.
+The only way I support Hillary was if Elizabeth Warren ran or Karl Marx was running #2016 #Clinton2016
+@HomeOfUncleSam @ScotsFyre @RWNutjob1 @SA_Hartdegen She's too old to understand the internet...that she can be fact checked.
+Because Communist Breadlines are not my thing! #NoHillary #WhyImNotVotingForHillary
+"@HillaryClinton bad wife, bad role model for women, bad lawyer, bad First Lady, bad Senator, horrible Secretary of State."
+"Everything Hillary touches ends up being a scam, a lie, a cover-up, or a failure. William L. Just who we want as president."
+Yes HRC subject 2 dbl standard Smh Come on @billclinton @HillaryClinton U Knew @ClintonFdn Donations Would b Scrutinized; Spun!
+#Hillary to stop for #pizza today to garner the #Italian vote. #MSM is worthless. #libertynothillary #HillNo
+I want America to great again #WhyImNotVotingForHillary
+"March 8, 2016 Ohio is holding our Primaries! The date is subject to change. #Ohio #OurChampion"
+@RIGHTZONE @WethePeoplePets Let's hope the VOTERS remember! #HillNo
+Hillary Clinton has not driven a car since 1996. #clintonfakerealityshow
+"@NaughtyBeyotch @TheRealMadman23  Don't care for #Fiorina, but it seems she taking the #sexist gut punches by the #media. #MSM"
+"@FoxNews @marthamaccallum @BillHemmer whose the opportunist now, #NoHillary2016"
+@HillaryClinton @WomenintheWorld we need to re-establish a #global system dominated by love and affection have #moral_humane RT
+#Hillary is as transparent as a brick wall #LibertyNotHillary
+@WSJ . Clinton Foundation to keep accepting bribes from foreign governments #WhyImNotVotingForHillary
+"@josephbenning I agree, these are better than what you had before, like a severe cold is better than pneumonia. Good luck."
+"What are you afraid of @HillaryClinton? If you can't answer questions from the press, why do we want you as POTUS"
+"Sorry, Hillary's new normal folk image doesn't take away from Behgnazi & her 0 foreign policy successes as Secretary of State."
+CEO pay the target for 2016 election. From someone who makes more than most CEO's but you drank the Kool-Aid.

parse_data.py

@@ -0,0 +1,16 @@
+import pandas as pd
+
+
+def write_tweets_to_txt(file):
+    df = pd.read_csv(file, encoding='ISO-8859-1')
+    column_data = df["Tweet"]
+    row_count = 0
+
+    with open('atheism_without_none.txt', 'w', encoding='ISO-8859-1') as f:
+        for value in column_data:
+            if(row_count % 43 == 0):
+                f.write('********************************************' + '\n')
+            f.write(str(value) + '\n')
+            row_count += 1
+
+write_tweets_to_txt('atheism_without_none.csv')

testSVM.py

@@ -0,0 +1,61 @@
+import numpy as np
+from nltk.tokenize import word_tokenize
+from snowballstemmer import TurkishStemmer
+import string
+from sklearn.feature_extraction.text import TfidfVectorizer
+
+import pandas as pd
+
+#stemmer = TurkishStemmer()
+
+def read_turkish_tweets(file):
+    df = pd.read_csv(file, encoding='windows-1254')
+    tweets = df["Tweet"].tolist()
+    targets = df["Target"].tolist()
+    return tweets, targets
+
+def detect_stopwords():
+    stopwords_df = pd.read_csv('turkish', header=None)
+    stop_words = stopwords_df[0].tolist()
+    #stop_words = stopwords.words('turkish')
+    stop_words.extend(string.punctuation)
+    stop_words.extend(["vs.", "vb.", "a", "i", "e", "rt", "#semst", "semst"])
+    stop_words = set(stop_words)
+    return stop_words
+
+
+def tokenize_tweet(tweet):
+    # Tokenization
+    tokens = word_tokenize(tweet)
+    stop_words = detect_stopwords()
+    normalized_tokens = [token.lower() for token in tokens]
+    filtered_tokens = [token for token in normalized_tokens if (token not in stop_words and not token.startswith("http"))]
+    #stemmed_tokens = [stemmer.stemWord(token) for token in filtered_tokens]
+
+    return filtered_tokens
+
+def extract_features_tfidf(tweets):
+    # Tokenization and preprocessing
+    tokenized_tweets = [tokenize_tweet(tweet) for tweet in tweets]
+
+    # Feature extraction: n-grams
+    tfidf_vectorizer = TfidfVectorizer(ngram_range=(1, 3), analyzer='word')
+    tfidf_features = tfidf_vectorizer.fit_transform([' '.join(tokens) for tokens in tokenized_tweets])
+
+    # Feature extraction: character n-grams
+    char_tfidf_vectorizer = TfidfVectorizer(ngram_range=(2, 5), analyzer='char')
+    char_tfidf_features = char_tfidf_vectorizer.fit_transform([' '.join(tokens) for tokens in tokenized_tweets])
+
+    # Feature extraction: sentiment lexicon features, target presence/absence, POS tags, encodings
+    # These features remain the same as before
+
+    # Combine all features
+    all_features = np.concatenate((tfidf_features.toarray(), char_tfidf_features.toarray()), axis=1)
+    np.savetxt('feature_matrix.csv', all_features, delimiter=',')
+    return all_features
+
+train_tweets, train_targets = read_turkish_tweets('translated_train_without_none.csv')
+
+print(extract_features_tfidf(train_tweets))
+
+#print(tokenize_tweet(train_tweets[0]))