From 7cf0a218286016f9e3066ffbd66ca6420d1545e9 Mon Sep 17 00:00:00 2001
From: Em-Vi <musthumuth1107@gmail.com>
Date: Sun, 19 Oct 2025 20:03:33 +0530
Subject: [PATCH] Add: [AI-ML] Sentiment-Analyzer-Using-Neural-Networks

---
 .../README.md                                 | Bin 0 -> 6462 bytes
 .../dataset.csv                               |  41 ++++++++++++
 .../main.py                                   |  63 ++++++++++++++++++
 .../requirements.txt                          | Bin 0 -> 80 bytes
 .../AI-ML/MiniProjects/faceBlurTool/README.md | Bin 1232 -> 1234 bytes
 5 files changed, 104 insertions(+)
 create mode 100644 Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/README.md
 create mode 100644 Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/dataset.csv
 create mode 100644 Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/main.py
 create mode 100644 Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/requirements.txt

diff --git a/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/README.md b/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..ca141e463f2003442e8d85b424a88bab0a6b08a3
GIT binary patch
literal 6462
zcmb`L&u&{q6vk(bgt}qF?v_B6R=AY@k*Z2m3T-MXHK<D5cG)<NlbR&9iJgX4UGNO-
zcn4mA4Ns8;AS71s5QXnIpC>c-#!ec9qFmoQbIzGF=kGW3&)X~EO1Kw>VGt%^KlH;!
zIMCZr?{_qFB~)QK?1ff%s=4>Wd+~oEENJ}aOI3Iv%Ax2tLMQC%uL`T77ha0C7Y_Bc
zre7rYCBtu3Skeg8ufl7M?H@~RhZmw7i)*COVfaR~UGY|u)QRH<n%UI*L^GQjN6$+7
zj`dwhYESR`QMSJ(;$$Z~n(b&Fp1n8%{YY<z;@J;d;zReB`eH?R`VH-tG(V2}7>O4C
zrSMQssE3hzBK^^=l8jxwjU)Y*czSxPv`W)fdy&2?zDg@*yFu+8<AKNGV`Z#}hKzkG
z3Vy%V_jZ(0pJNA^0*~4*_9zzX%O2ZPniWYv3&z;lBRxCf7)+_)y7~@&+7uu59Z0@c
z#8d1B+aNjlV@npo`jwtrdg2}Hi&Z}hE7GA72Rd3S{ABGvsIVmwvL_nHLq~5!O;1lW
z#}ECeNs1EmLO=X$-H>Z5U`?z~R2~$(jHkG<^i6RPA(owO(o%-Bjx9ufEQt4uqwjww
zN3iB3#wONCSwwE#n;9u%Sql#mOQ%OoH(DX(7M5m5&WjuRA~Tx#*Nxf`H;!|3@=nm`
zN#sEi*8Wno+u9jkYK3Q!V^b9NMBS$3BKizslv;cI|6SZi)@vY3pfNdyE_ft)G3(Ke
zUW5x>@GomVl6)e9_=4N@W?QuEf~aPFYj-jW$%!)d$CA)jbs)D|(j!HE)?!Jcj#RX8
zrP+@f*~aKM(m0-SjQ6DrQGKYHS>7x2-1E@yN(XW~Me+IK?C#MYzg3|ve%F#U=|V>D
z)M5>*JXY03&txTH5w9~x=1?Q4JhtEu7qe$7ad{qH$;*;=o+YcWC}}%|Z7&zH>+wU@
zBH7Y&=H(b_*5{n%r7q`*5q3B`N9q;aIg=_1F`KdnYwjJ}d0rj(<<~0Q*Z%Me>rIY1
z+TzFWDFZx<-PRjpSLM;MqGUZ>7QHhKdDx5Sycw?Pd3wDlqYs`#akB#=u%j+ceROX<
zX_c`ML8@JE#IY*M-07E<=N(xhb;n8+=letEnq)TX)tk5~v$pR4^{m2dnT)efW<OFV
zN|oiq3X9=)*Q|_622;z?>FFXH(TW$v-LwEQoBbS`a>X1GJ$Vy*;ERlc$cKIXfz^ye
zs71ujE7kGlnylDF3C$HBDh9HA_u`?l7K!W_zxE>L7-%KpCFST?{-)~mMYB6yJ*}8g
zhppf}V%fYyX9Qgf+m5vka_9)gq5Lv}PF<}8f}{WA1@@dP=q&7!w9FNx*%Rt{&06#j
z^6aeM{^O+!Z!2LHV;yT<uSe{3J^oqlix^cS$sb8?XD#{b{kCG?AV0rVKcS0&rJN0D
zGl_jP%jC^pXYs6SH6sd8E;eIrEODT(UXj(-7O7<~;Co(uw+8U>#{cXJc=zZpb*Qi8
zGccdy&1hmrvJ<ay(}E~I4;SA4S>>8Z*09eex=Wm^n(JscMA7`ZnR^OxOCOs=4t`7I
zME7fAL@M%aRoG8decc`14c$A5Fk=#<>cw!qM(e7Mb|wB;FIIW3cguJ#NlCj!t+rmo
zVz$ursfF6&GQxVQxAOFIG8*gHj#J-(B4MJlU7a&y?ZqhqOgKFwKzfid(0|2zWjA;B
zgCMCcXrAy?H9ZTISb?ksUHP0yT>$$fj=h=ox-qraN~{|`2V{)tN|B<8vmQP^50YqB
z7HpX6T8<@GFwj@hmft4wakMalMKUf;=O=QQOql48b=?2S4=N>NP?#H8^^RKBwj574
zr}kNn9jty1N_I(gJNv>t5v<u&EU}Mn>=~}-)GsP3GuEk$emqTIOL3nu5GQdu3c1{r
zO?D7*NSvP?x2C&sUD#DsYD*&eq)Jjp*v%a#qz}$=tPvpz>7yC@CF6YxJ&C%|FQKFI
z9GP~VZD<tcE>J&ZVjJ7&q!-WXPj`?0R>f_ivelUJX}7NTWq0~mI~x^e1nx_WFt5gO
zILi!Lk{wFV-x2@qa7!6@E!IEimt*`nAF$l%wnD<Q;~HRkJYeLUPJQ|2<YMd|?heVP
zj<q#;pdR_iZlW(Vv>lRF=5;X6SWhI8p2-;yjrely$iB+<NfJRsoXO~4#zOO-&{CO=
zcShoxPiq9bDr)1v(<slL%NX?C5cWL$Je&J`)U#!z(GoAGQxzB|b<J6O#jft(cq&!g
zN_S1wv{#pv2x`X^U5Y58zYLYp_?|mGYH#xVto8C(w(IE~x8$hi-n{q*Gu>ptlOx93
zSy!UXxPUXR?_4<F`Sv2ORe#I6zFY50tDkmSE-M|nK9Ux`UBsro5!_UTK3U^@E9#!+
zvo|;v{0wK-wONJ8w$vla{jvJ2{)SCCllw+0aX`Hibxj?NRkt<5Eh!NMf_$xCBEdbK
z(-84Nf1*Y?i<7R@1IKWx4DP?&6#k?ef(NS3>7Edj%xRi?PW+07j>E*8Z|jU!+Tw4C
z>%M-CP>E0Kl(BKbaA!rKwddr8P2dK}a1v#w?&bKcr(JK$A1M#=tyx<foE*{OzUW-{
zvfrII)^Snsu%_<3B6_Sw<S*-Pp&f5Ik}i0Y-8-haqvebUdYxU3vo1@b01eSeSP{K9
f^!`MavIZq4p-R)aIq~{@khFTBZ?McxfiC|Amw62t

literal 0
HcmV?d00001

diff --git a/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/dataset.csv b/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/dataset.csv
new file mode 100644
index 00000000..55a4c10f
--- /dev/null
+++ b/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/dataset.csv
@@ -0,0 +1,41 @@
+review,sentiment
+"What a fantastic comeback by Liverpool!",positive
+"The referee ruined the match completely.",negative
+"It was an okay game, nothing special.",neutral
+"Messi’s goal was unbelievable!",positive
+"The defense was terrible today.",negative
+"The match ended in a draw, fair enough.",neutral
+"Ronaldo carried the whole team again!",positive
+"Our midfield looked clueless tonight.",negative
+"Good tactical game from both sides.",positive
+"Not the most exciting match to watch.",neutral
+"The keeper made some world-class saves!",positive
+"We wasted too many chances up front.",negative
+"Both teams deserved the point.",neutral
+"Brilliant performance by the youngsters!",positive
+"The coach has no idea what he's doing.",negative
+"The atmosphere in the stadium was electric!",positive
+"It was a boring goalless draw.",neutral
+"Terrible refereeing cost us the game.",negative
+"The fans were amazing throughout the match.",positive
+"Nothing went our way today.",negative
+"The new signing looked promising.",positive
+"The first half was dull but improved later.",neutral
+"We dominated possession but couldn’t score.",negative
+"Superb hat-trick from Haaland!",positive
+"The team lacked motivation and energy.",negative
+"The substitution made a huge impact.",positive
+"VAR decisions were so inconsistent again.",negative
+"Overall, a decent performance from both sides.",neutral
+"Best match of the season so far!",positive
+"That red card completely changed the game.",negative
+"Fair result, both teams played well.",neutral
+"Insane dribbling and pace from Mbappe!",positive
+"The defense collapsed after halftime.",negative
+"Not bad, but expected better from them.",neutral
+"Unbelievable last-minute winner!",positive
+"Sloppy passing all over the field.",negative
+"The crowd was quiet most of the time.",neutral
+"Fantastic teamwork and creativity today!",positive
+"Poor finishing cost us again.",negative
+"It was a balanced game overall.",neutral
diff --git a/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/main.py b/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/main.py
new file mode 100644
index 00000000..2789f1f7
--- /dev/null
+++ b/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/main.py
@@ -0,0 +1,63 @@
+import pandas as pd # for data manipulation
+from sklearn.model_selection import train_test_split # for splitting data
+# from sklearn.feature_extraction.text import CountVectorizer # for turning text into vectors
+# from sklearn.linear_model import LogisticRegression # for classification on data
+# from sklearn.metrics import accuracy_score # for evaluating model performance
+from sklearn.feature_extraction.text import TfidfVectorizer # checks how important a word is to a document in a collection 
+from sklearn.preprocessing import LabelEncoder # converts nuetral, positive and negative to values
+from tensorflow.keras.models import Sequential # for creating neural network model
+from tensorflow.keras.layers import Dense, Dropout # dense helps to build a neural network dropout enseures no overfitting
+
+# Load dataset
+data = pd.read_csv("dataset.csv")
+
+# train-test split:
+# splitting 80-20 for training and testing. this ensures that model would be able to genralize well on unseen data.
+# random state locks the data on randomizing everytime we run the code
+X_train, X_test,y_train, y_test = train_test_split(data['review'], data["sentiment"],test_size=0.2, random_state=42)
+
+# text vecotrization:
+# converts text data into numerical vectors(row, column format) that machine learning models can understand.
+# the tfidf was upgraded and max_featues keeps 5000 top most relevant words
+# vectorizer = CountVectorizer()
+vectorizer = TfidfVectorizer(max_features=5000)
+
+# fit transform learns vocabulary while transform uses same vocabulary data
+X_train_vectors = vectorizer.fit_transform(X_train)
+X_test_vectors = vectorizer.transform(X_test)
+
+# Update encode labels
+encoder = LabelEncoder()
+y_train_enc = encoder.fit_transform(y_train)
+y_test_enc = encoder.transform(y_test)
+
+# Model training:
+# model = LogisticRegression()
+# # learning from training data X_train_vectors the vectored sentence and y_train the labels
+# model.fit(X_train_vectors, y_train)
+# upgrade building a neural network
+model = Sequential([
+    Dense(128, activation='relu', input_shape=(X_train_vectors.shape[1],)),
+    Dropout(0.3),
+    Dense(64, activation='relu'),
+    Dense(3, activation='softmax')  # 3 classes: positive, negative, neutral
+])
+
+#compile the model
+model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
+
+#Train model
+model.fit(X_train_vectors.toarray(),y_train_enc, epochs=10, batch_size=32, validation_split=0.2)
+
+# Evaluation
+# predictions = model.predict(X_test_vectors)
+# accuracy = accuracy_score(y_test, predictions)
+loss, accuracy = model.evaluate(X_test_vectors.toarray(), y_test_enc)
+print(f"Model Accuracy: {accuracy*100:.2f}%")
+
+# Prediction:
+user_input = input("Enter a football match review: ")
+user_vector = vectorizer.transform([user_input]).toarray()
+prediction = model.predict(user_vector)
+sentiment = encoder.inverse_transform([prediction.argmax()])[0]
+print(f"Predicted Sentiment: {sentiment.capitalize()}")
\ No newline at end of file
diff --git a/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/requirements.txt b/Domains/AI-ML/MiniProjects/SentimentAnalyzerUsingNeuralNetworks/requirements.txt
new file mode 100644
index 0000000000000000000000000000000000000000..6e6cc779fc75abb7e8c28cc7a17b42a41539c9d6
GIT binary patch
literal 80
zcmW-YOA3H63<YN`coYvMen8Z=v=#C4>gXaN^D^(b6Fq^^U1({%4|WpQPlbVn)v0uB
SOnO=8GpiJAMU&ZI<m3gv_z(;L

literal 0
HcmV?d00001

diff --git a/Domains/AI-ML/MiniProjects/faceBlurTool/README.md b/Domains/AI-ML/MiniProjects/faceBlurTool/README.md
index 159934f255fa0e92f1f3b3c6446f66cbcb2c18dd..0cc9abdc8df8daa119f1d7461f93e793adc93d04 100644
GIT binary patch
delta 12
Tcmcb>d5Lp^4rAs<?WHUL9}Way

delta 10
Rcmcb_d4Y3+&PKf@EC3iJ1NZ;{