fix: Improve handling of stopwords comparison without nltk

app.py

@@ -2,15 +2,13 @@ from typing import List
 
 import string
 import re
+import requests
 import tweepy
 import hdbscan
-import nltk
+
 import numpy as np
 import streamlit as st
 
-nltk.download()
-
-from nltk.corpus import stopwords
 from gensim.utils import deaccent
 from bokeh.models import ColumnDataSource, HoverTool, Label
 from bokeh.palettes import Colorblind as Pallete
@@ -27,6 +25,10 @@ model_to_use = {
     "Use all the ones you know (~15 lang)": "paraphrase-multilingual-MiniLM-L12-v2"
 }
 
+
+stopwords_list = requests.get("https://gist.githubusercontent.com/rg089/35e00abf8941d72d419224cfd5b5925d/raw/12d899b70156fd0041fa9778d657330b024b959c/stopwords.txt").content
+stopwords = set(stopwords_list.decode().splitlines()) 
+
 def _remove_unk_chars(txt_list: List[str]):
     txt_list = [re.sub('\s+', ' ', tweet) for tweet in txt_list]
     txt_list = [re.sub("\'", "", tweet) for tweet in txt_list]
@@ -50,9 +52,8 @@ def _remove_punctuation(txt_list: List[str]):
     return [' '.join([word.translate(str.maketrans('', '', punctuation)) for word in tweet]) for tweet in txt_list]
 
 def _remove_stopwords(txt_list: List[str]):
-    stopwords_str = stopwords.words('english')
     txt_list = [tweet.split(' ') for tweet in txt_list]
-    return [' '.join([word for word in tweet if word not in stopwords_str]) for tweet in txt_list]
+    return [' '.join([word for word in tweet if word not in stopwords]) for tweet in txt_list]
 
 preprocess_pipeline = [
     _remove_unk_chars,