Update app.py

app.py

@@ -54,7 +54,7 @@ if page == 'Home Page':
     with st.expander("Conclusion"): # conclusion
         st.caption(
             '''
-            The dataset is well maintained and has a balance target class. EDA shows that most people are quite concise when it comes to expressing their feelings, only a few whom can write really long and be more open about what they feel. Life, good, day, something, going and well are words that are most often expressed when people describe a joyful experience. Would, even, way, think, and angry are words that's the most often expressed when a person is mad. Still, bit, strange, ank think are words that is most often expressed when a person describing a fearful experience. Anger and fear are a special case. Anger words such as 'would' and 'even' is an auxillary verb that indicates a mood or tense. Fear is a broad emotion where it can classify into fear of the unknown that involes anxiety and fear as if being scared or teriffied, giving goosebumps. The best model uses LSTM network with 93% accuracy. There still seems to be a lot of noise therefor, focusing on preprocessing such as adding more stopwords could help the model learn patterns more accurately in the future.
+            The dataset is well maintained and has a balance target class. EDA shows that most people are quite concise when it comes to expressing their feelings, only a few whom can write really long and be more open about what they feel. Life, good, day, something, going and well are words that are most often expressed when people describe a joyful experience. Would, even, way, think, and angry are words that's the most often expressed when a person is mad. Still, bit, strange, ank think are words that is most often expressed when a person describing a fearful experience. Anger and fear are a special case. Anger words such as 'would' and 'even' is an auxillary verb that indicates a mood or tense. Fear is a broad emotion where it can classify into fear of the unknown that involes anxiety and fear as if being scared or teriffied, giving goosebumps. The best model uses LSTM network with 93% accuracy. There still seems to be a lot of noise therefor, focusing on preprocessing such as adding more stopwords could help the model learn patterns more accurately in the future. The dataset currently includes just three emotions, but incorporating data on a wider range of emotions would enhance the model's ability to classify more diverse emotional states.
             '''
         )