app.py

from bertopic import BERTopic
import streamlit as st
import streamlit.components.v1 as components
from datasets import load_dataset
import pandas as pd
from sentence_transformers import SentenceTransformer
from umap import UMAP
from hdbscan import HDBSCAN
from sklearn.feature_extraction.text import CountVectorizer

st.set_page_config(page_title='eRupt Topic Trendy (e-Commerce x Social Media)', page_icon=None, layout='centered', initial_sidebar_state='auto')

st.markdown("<h1 style='text-align: center;'>Topic Trendy</h1>", unsafe_allow_html=True)
#BerTopic_model = BERTopic.load("my_topics_model")

#sentence_model = SentenceTransformer("all-MiniLM-L6-v2")
#umap_model = UMAP(n_neighbors=15, n_components=2, min_dist=0.1, metric="cosine")
#hdbscan_model = HDBSCAN(min_cluster_size=5, min_samples = 3, metric="euclidean", prediction_data=True)
#vectorizer_model = CountVectorizer(lowercase = True, ngram_range=(1, 3), analyzer="word", max_df=1.0, min_df=0.5, stop_words="english")

#kw_model = BERTopic(embedding_model=sentence_model, umap_model = umap_model, hdbscan_model = hdbscan_model, vectorizer_model = vectorizer_model, nr_topics = "auto", calculate_probabilities = True)
#BerTopic_model = kw_model

topic = pd.read_csv('./Data/tiktok_utf8.csv')

timestamps = topic.date.to_list()
tiktok = topic.text.to_list()

vectorizer_model = CountVectorizer(stop_words="english")
topic_model = BERTopic(verbose=True,vectorizer_model=vectorizer_model)

@st.cache()
def fit_transform(model, docs):
    topics, probs = model.fit_transform(docs)
    return topics, probs

topics, probs = fit_transform(topic_model, tiktok)

#topics_over_times = topic_model.topics_over_time(tiktok, topics, timestamps, nr_bins=20)
#topic_model.visualize_topics_over_time(topics_over_times, top_n_topics=30)
#topics, probs = topic_model.fit_transform(tiktok)

placeholder = st.empty()
text_input = placeholder.text_area("Enter product topic here", height=300)

similar_topics, similarity = topic_model.find_topics(text_input, top_n=20)

most_similar = similar_topics[0]
print(similar_topics[0])
print("Most Similar Topic Info: \n{}".format(topic_model.get_topic(most_similar)))
print("Similarity Score: {}".format(similarity[0]))

answer_as_string = topic_model.get_topic(most_similar)

st.text_area("Most Similar Topic List is Here",answer_as_string,key="topic_list")
st.image('https://freepngimg.com/download/keyboard/6-2-keyboard-png-file.png',use_column_width=True)
st.markdown("<h6 style='text-align: center; color: #808080;'>Created By LiHE</a></h6>", unsafe_allow_html=True)