Spaces:
No application file
No application file
File size: 9,288 Bytes
21b78eb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 |
import streamlit as st
import streamlit.components.v1 as components
# from PIL import Image
import requests
import pickle as pkl
import re
import pandas as pd
import json
import PyPDF2
import os
from examples import EXAMPLES
from ner_examples import NER_EXAMPLES
from annotated_text import annotated_text
from PIL import Image
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator
import matplotlib.pyplot as plt
import numpy as np
try:
import StringIO
except ImportError:
from io import StringIO
def get_topics(url, data):
r = requests.post(url, json = data)
print(r.status_code)
if r.status_code != 200:
return None
else:
return r.json()
def get_ner(url, data):
r_ner = requests.post(url, json = data)
print(r_ner.status_code)
if r_ner.status_code != 200:
return None
else:
return r_ner.json()
# def download_link(
# content, label="Download", filename="file.txt", mimetype="text/plain"
# a fuction used for word cloud
def plot_wordcloud(list_topic:list) -> None:
'''
Plot a wordcloud of top 20 words from the input text
masked by world logo
'''
text = " ".join(word for word in list_topic)
# Create stopword
stopwords = set(STOPWORDS)
# mask = np.array(Image.open('/home/ked/client/img/worldmap1.png'))
wordcloud = WordCloud(
background_color='white',
stopwords=stopwords,
random_state=42,
max_words=20,
max_font_size=80).generate(text)
# plt.figure(figsize=(10,10))
# plt.imshow(wordcloud, interpolation="bilinear")
# plt.axis("off")
# plt.show()
fig, ax = plt.subplots(figsize = (10, 10))
ax.imshow(wordcloud, interpolation = "bilinear")
plt.axis("off")
st.pyplot(fig)
url = 'http://backendtopic:8000/topic/predict'
url_ner = 'http://backendner:9000/ner/predict'
st.set_page_config(page_title="GlgCapstone-Demo", page_icon=":star:", layout="wide")
st.subheader("GLG Topic Modeling and Named Entity Recognation")
tab1, tab2 = st.tabs(["Topic Analysis", "Named Entity Recognition"])
with tab1:
with st.expander("ℹ️ About Topic Model", expanded=True):
st.write(
"""
- Topic model is a type of statistical model for discovering the abstract "topics" that occur in a collection of documents.
- Topic models can help to organize and offer insights for us to understand large collections of unstructured text bodies.
"""
)
st.markdown("")
# st.set_page_config(layout="wide")
# st.markdown("""
# <style>
# .big-font {
# font-size:300px !important;
# }
# </style>
# """, unsafe_allow_html=True)
# st.markdown('<p class="big-font">Hello World !!</p>', unsafe_allow_html=True)
# with st.form(key="my_form"):
col1, col2, col3 = st.columns([3,1,3])
with col1:
prompts = list(EXAMPLES.keys()) + ["Select a document"]
prompt = st.selectbox(
'Example Inputs',
prompts,
index=2
)
if prompt == "Select a document":
prompt_box = ""
else:
prompt_box = EXAMPLES[prompt]
with col3:
uploaded_file = col3.file_uploader("Upload pdf document", type=".pdf")
if uploaded_file:
# creating a pdf file object
# pdfFileObj = StringIO(uploaded_file.getvalue().decode("utf-8"))
# creating a pdf reader object
pdfReader = PyPDF2.PdfFileReader(uploaded_file)
# printing number of pages in pdf file
print(pdfReader.numPages)
# creating a page object
pageObj = pdfReader.getPage(0)
prompt_box = pageObj.extractText()
# closing the pdf file object
# pdfFileObj.close()
doc_txt = st.text_area(
"Document:",
prompt_box, height=200
)
submit_button = st.button(label="Generate topics")
# if not doc_txt:
# st.stop() # pop up message
if submit_button:
if doc_txt != "" and len(doc_txt.split(" ")) > 12:
with st.spinner("Generating topics..."):
data = {"document": {"0": doc_txt}}
topics = get_topics(url, data)
st.markdown("Model Output")
tab1_result, tab2_result = st.tabs(["Result Tables", "Result Wordcloud" ])
st.header("")
df_global = pd.DataFrame(topics['topics']['0']['global'].items())
df_global['label'], df_global['topics'] = df_global[1].apply(lambda x: x['labels']), df_global[1].apply(lambda x: x['topics'])
df_global = df_global.set_index(df_global[0])
df_global.drop(1, axis=1, inplace=True)
df_local = pd.DataFrame(topics['topics']['0']['local'].items())
df_local['label'], df_local['topics'] = df_local[1].apply(lambda x: x['labels']), df_local[1].apply(lambda x: x['topics'])
df_local = df_local.set_index(df_local[0])
df_local.drop(1, axis=1, inplace=True)
with tab1_result:
st.header("Global Topics")
st.table(df_global)
st.header("Local Topics")
st.table(df_local)
with tab2_result:
global_topics = df_global['topics'].tolist()
global_labels = df_global['label'].tolist()
local_topics = df_local['topics'].tolist()
local_labels = df_local['label'].tolist()
global_topic_label = global_topics + global_labels
local_topic_label = local_topics + local_labels
col4, col_, col5 = st.columns([2,1,2])
with col4:
st.header("Global Topics as a wordcloud")
plot_wordcloud(global_topic_label)
with col5:
st.header("Local Topics as a wordcloud")
plot_wordcloud(local_topic_label)
else:
st.warning('Please insert a document', icon="⚠️")
with tab2:
with st.expander("ℹ️ Named Entity Recognition", expanded=True):
st.write(
"""
Named Entity Recognition is the task of identifying named entities (people, locations, organizations, etc.) in the input text.
"""
)
tab3, tab4 = st.tabs(["Demo", "Model Info"])
with tab3:
prompts_ner = list(NER_EXAMPLES.keys()) + ["Select a Sentence"]
prompt_ner = st.selectbox(
'Example Document',
prompts_ner,
index=3
)
if prompt_ner == "Select a Sentence":
prompt_box = ""
else:
prompt_box = NER_EXAMPLES[prompt_ner]
sent_txt = st.text_area(
"Sentence:",
prompt_box, height=100)
submit_button_ner = st.button(label="Run Model")
if submit_button_ner:
if sent_txt != "":
with st.spinner("Generating entities..."):
sent_data = {"sentence": sent_txt}
ner_output = get_ner(url_ner, sent_data)
st.markdown("Model Output")
st.markdown("Entities")
tokens_ner = ner_output['ner_tags']['tokens'][1:-1]
labels_ner = ner_output['ner_tags']['labels'][1:-1]
print(zip(tokens_ner, labels_ner))
annotated_list = []
ner_entities = ['per','gpe','geo','art','eve','org','tim','nat']
for i,token_label in enumerate(zip(tokens_ner, labels_ner)):
token, label = token_label[0], token_label[1]
if label.lower() not in ['o', 'pad']:
tag = label.split("-")
if tag[0] == "B":
collector = token
flag = True
j = i+1
while flag:
if labels_ner[j].lower() not in ['o', 'pad']:
if labels_ner[j].split("-")[1] != tag[1]:
flag = False
else:
collector = collector + " " +tokens_ner[j]
j += 1
else:
flag = False
annotated_list.append((collector, tag[1]))
else:
annotated_list.append(token+" ")
print(annotated_list)
# st.write(annotated_list)
annotated_text(*annotated_list)
with tab4:
data_path = os.getcwd()
os.path.join(data_path,"data/modelcard.csv")
df = pd.read_csv(os.path.join(data_path,"data/modelcard.csv"), sep=',')
# df = df.rename(columns={'0':'','1':''})
st.table(df)
|