Commit
Browse files- Pages/About.py +20 -19
- Pages/Models.py +11 -1
Pages/About.py
CHANGED
|
@@ -1,46 +1,47 @@
|
|
| 1 |
import streamlit as st
|
| 2 |
|
| 3 |
|
| 4 |
-
# Display the about page of the app with information about the creator, code
|
| 5 |
def about_page():
|
| 6 |
-
st.title("About Us")
|
| 7 |
with st.container():
|
| 8 |
-
col = st.columns([1, 1])
|
| 9 |
with col[0]:
|
|
|
|
| 10 |
st.write("\n")
|
| 11 |
st.write("\n")
|
| 12 |
-
st.
|
| 13 |
-
|
| 14 |
-
"This app was created by [Harshit Singh](https://harsh502s.github.io), Poorvi Singh and Samruddhi Raskar as a part of their MSc Data Science 3rd semester project."
|
| 15 |
)
|
| 16 |
st.write("\n")
|
| 17 |
-
st.
|
|
|
|
|
|
|
|
|
|
| 18 |
st.write("\n")
|
| 19 |
-
st.
|
| 20 |
-
"The data on which these models are trained can be found [here](https://www.kaggle.com/datasets/harsh502s/stackexchange-tag-dataset)."
|
|
|
|
| 21 |
)
|
| 22 |
with col[1]:
|
| 23 |
-
st.image("Group.svg", width=
|
| 24 |
|
| 25 |
st.write("\n")
|
| 26 |
st.write("\n")
|
| 27 |
|
| 28 |
with st.container():
|
| 29 |
-
col = st.columns([1, 2])
|
| 30 |
with col[0]:
|
| 31 |
st.image("Robot.svg", width=350)
|
| 32 |
with col[1]:
|
| 33 |
st.title("Models Used:")
|
| 34 |
-
st.
|
| 35 |
-
"""1.
|
| 36 |
-
is a topic modeling technique that leverages BERT embeddings and c-TF-IDF to create dense clusters allowing for easily interpretable topics whilst keeping important words in the topic descriptions."""
|
| 37 |
)
|
| 38 |
-
st.
|
| 39 |
-
"""2.
|
| 40 |
-
is a minimal and easy-to-use keyword extraction technique that leverages BERT embeddings to create keywords and keyphrases that are most similar to a document."""
|
| 41 |
)
|
| 42 |
-
st.
|
| 43 |
-
"""3. Convolutional Neural Networks (CNNs) are used for text classification. CNNs can identify patterns in text data, such as bigrams, trigrams, or n-grams. CNNs are translation invariant, so they can detect these patterns regardless of their position in the sentence."""
|
| 44 |
)
|
| 45 |
|
| 46 |
|
|
|
|
| 1 |
import streamlit as st
|
| 2 |
|
| 3 |
|
| 4 |
+
# Display the about page of the app with information about the creator, code and data.
|
| 5 |
def about_page():
|
|
|
|
| 6 |
with st.container():
|
| 7 |
+
col = st.columns([1.5, 1])
|
| 8 |
with col[0]:
|
| 9 |
+
st.title("About Us")
|
| 10 |
st.write("\n")
|
| 11 |
st.write("\n")
|
| 12 |
+
st.markdown(
|
| 13 |
+
"""##### This app was created by [Harshit Singh](https://harsh502s.github.io), Poorvi Singh and Samriddhi Raskar as a part of their MSc Data Science 3rd semester project."""
|
|
|
|
| 14 |
)
|
| 15 |
st.write("\n")
|
| 16 |
+
st.markdown(
|
| 17 |
+
"""##### The code for this app can be found [here](https://github.com/Harsh502s/Autonomous-Text-Tagging-System)""",
|
| 18 |
+
unsafe_allow_html=True,
|
| 19 |
+
)
|
| 20 |
st.write("\n")
|
| 21 |
+
st.markdown(
|
| 22 |
+
"""##### The data on which these models are trained can be found [here](https://www.kaggle.com/datasets/harsh502s/stackexchange-tag-dataset/data).""",
|
| 23 |
+
unsafe_allow_html=True,
|
| 24 |
)
|
| 25 |
with col[1]:
|
| 26 |
+
st.image("Group.svg", width=325)
|
| 27 |
|
| 28 |
st.write("\n")
|
| 29 |
st.write("\n")
|
| 30 |
|
| 31 |
with st.container():
|
| 32 |
+
col = st.columns([1.5, 2])
|
| 33 |
with col[0]:
|
| 34 |
st.image("Robot.svg", width=350)
|
| 35 |
with col[1]:
|
| 36 |
st.title("Models Used:")
|
| 37 |
+
st.markdown(
|
| 38 |
+
"""###### 1. BERTopic is a topic modeling technique that leverages BERT embeddings and c-TF-IDF to create dense clusters allowing for easily interpretable topics whilst keeping important words in the topic descriptions."""
|
|
|
|
| 39 |
)
|
| 40 |
+
st.markdown(
|
| 41 |
+
"""###### 2. KeyBERT is a minimal and easy-to-use keyword extraction technique that leverages BERT embeddings to create keywords and keyphrases that are most similar to a document."""
|
|
|
|
| 42 |
)
|
| 43 |
+
st.markdown(
|
| 44 |
+
"""###### 3. Convolutional Neural Networks (CNNs) are used for text classification. CNNs can identify patterns in text data, such as bigrams, trigrams, or n-grams. CNNs are translation invariant, so they can detect these patterns regardless of their position in the sentence."""
|
| 45 |
)
|
| 46 |
|
| 47 |
|
Pages/Models.py
CHANGED
|
@@ -175,6 +175,16 @@ def semi_unsupervised_page_keybert():
|
|
| 175 |
|
| 176 |
# Display the model page of the app
|
| 177 |
def model_page():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 178 |
st.title("Select a model to use:")
|
| 179 |
with st.container():
|
| 180 |
tab1, tab2, tab3 = st.tabs(
|
|
@@ -187,7 +197,7 @@ def model_page():
|
|
| 187 |
with tab3:
|
| 188 |
unsupervised_page_bertopic()
|
| 189 |
with st.container():
|
| 190 |
-
with st.expander("Example Texts", expanded=
|
| 191 |
st.markdown(
|
| 192 |
"""
|
| 193 |
### Here are 5 examples of questions from Stack Exchange. Try them out!
|
|
|
|
| 175 |
|
| 176 |
# Display the model page of the app
|
| 177 |
def model_page():
|
| 178 |
+
stype_for_page = """
|
| 179 |
+
<style>
|
| 180 |
+
button.st-emotion-cache-c766yy.ef3psqc11:hover {
|
| 181 |
+
scale: 1.07;
|
| 182 |
+
transition-duration: 0.3s;
|
| 183 |
+
}
|
| 184 |
+
</style>
|
| 185 |
+
"""
|
| 186 |
+
st.markdown(stype_for_page, unsafe_allow_html=True)
|
| 187 |
+
|
| 188 |
st.title("Select a model to use:")
|
| 189 |
with st.container():
|
| 190 |
tab1, tab2, tab3 = st.tabs(
|
|
|
|
| 197 |
with tab3:
|
| 198 |
unsupervised_page_bertopic()
|
| 199 |
with st.container():
|
| 200 |
+
with st.expander("Example Texts", expanded=False):
|
| 201 |
st.markdown(
|
| 202 |
"""
|
| 203 |
### Here are 5 examples of questions from Stack Exchange. Try them out!
|