ringkas-ulas / app.py
kensvin's picture
Setup
8be6571
raw
history blame
4.48 kB
import re
import requests
import pandas as pd
import unicodedata as uni
import emoji
from langchain.chat_models import ChatOpenAI
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.document_loaders import DataFrameLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.vectorstores import FAISS
from langchain.chains import RetrievalQA
import gradio as gr
SHOPEE_API_URL = """https://shopee.co.id/api/v2/item/get_ratings?filter=0&flag=1&itemid={item_id}&limit=20&offset={offset}&shopid={shop_id}&type=0"""
shop_id = ""
item_id = ""
item = {}
LIMIT = 1000 # Limit to 1000 reviews so that processing does not take too long
def get_product_id(URL):
# Get shop id and item id from input URL
r = re.search(r"i\.(\d+)\.(\d+)", URL)
shop_id, item_id = r[1], r[2]
return shop_id, item_id
def scrape(URL):
try:
shop_id, item_id = get_product_id(URL)
except:
return None
offset = 0
reviews = []
while True:
# Get JSON data using shop_id and item_id from input URL
data = requests.get(
SHOPEE_API_URL.format(shop_id=shop_id, item_id=item_id, offset=offset)
).json()
i = 1
for i, review in enumerate(data["data"]["ratings"], 1):
reviews.append(review["comment"])
if i % 20:
break
offset += 20
if offset >= LIMIT:
break
df = pd.DataFrame(reviews, columns=["comment"])
return df
# Clean
def clean(df):
df = df.dropna().copy().reset_index(drop=True) # drop reviews with empty comments
df = df[df["comment"] != ""].reset_index(drop=True) # remove empty reviews
df["comment"] = df["comment"].apply(lambda x: clean_text(x)) # clean text
df = df[df["comment"] != ""].reset_index(drop=True) # remove empty reviews
return df
def clean_text(text):
text = uni.normalize("NFKD", text) # normalise characters
text = emoji.replace_emoji(text, "") # remove emoji
text = re.sub(r"(\w)\1{2,}", r"\1", text) # repeated chars
text = re.sub(r"[ ]+", " ", text).strip() # remove extra spaces
return text
# LLM
OpenAIModel = "gpt-3.5-turbo"
llm = ChatOpenAI(model=OpenAIModel, temperature=0.1)
# Embeddings
embeddings = HuggingFaceEmbeddings(model_name="Blaxzter/LaBSE-sentence-embeddings")
cache_URL = ""
db = None
qa = None
def generate(URL, query):
global cache_URL, db, qa
if URL != cache_URL:
# Get reviews
try:
reviews = scrape(URL)
# Clean reviews
cleaned_reviews = clean(reviews)
# Load data
loader = DataFrameLoader(cleaned_reviews, page_content_column="comment")
documents = loader.load()
except Exception as e:
return "Error getting reviews: " + str(e)
# Split text
text_splitter = RecursiveCharacterTextSplitter(
chunk_size=1000, chunk_overlap=50
)
docs = text_splitter.split_documents(documents)
cache_URL = URL
# Vector store
db = FAISS.from_documents(docs, embeddings)
# Chain to answer questions
qa = RetrievalQA.from_chain_type(llm=llm, retriever=db.as_retriever())
return qa.run(query)
# Gradio
product_box = gr.Textbox(
label="URL Produk", placeholder="URL produk dari Shopee Indonesia"
)
query_box = gr.Textbox(
lines=2,
label="Kueri",
placeholder="Contoh: Apa yang orang katakan tentang kualitas produknya?, Bagaimana pendapat orang yang kurang puas dengan produknya?",
)
gr.Interface(
fn=generate,
inputs=[product_box, query_box],
outputs=gr.Textbox(label="Jawaban"),
title="RingkasUlas",
description="Bot percakapan yang bisa meringkas ulasan-ulasan produk di Shopee Indonesia (https://shopee.co.id/). Harap bersabar, bot ini dapat memakan waktu agak lama saat mengambil ulasan dari Shopee dan menyiapkan jawabannya.",
allow_flagging="never",
examples=[
[
"https://shopee.co.id/Bantal-Selimut-Balmut-Mini-Karakter-kain-CVC-i.2392232.8965506?xptdk=324a77c0-7860-4059-b00d-5d3b340f8dfe",
"Apa yang orang katakan tentang kualitas produknya?",
],
[
"https://shopee.co.id/Bantal-Selimut-Balmut-Mini-Karakter-kain-CVC-i.2392232.8965506?xptdk=324a77c0-7860-4059-b00d-5d3b340f8dfe",
"Bagaimana pendapat orang yang kurang puas dengan produknya?",
],
],
).launch()