File size: 5,471 Bytes
42472b3 |
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 |
import streamlit as st
import openai
import sys, getopt
from datetime import datetime
from streamlit.components.v1 import html
import boto3
from main import chatgpt_prompt, get_chatgpt_resp, generate_kyc_output, gsearch, save_to_s3
# Function to perform the search
# This is a placeholder function, replace it with your actual search implementation
def perform_search(pname, keywords, num_results):
# record current timestamp
start_time = datetime.now()
# Google search for the person name and get the first 20 query links
query = pname + " " + keywords
search_links = gsearch(query, num_results)
# Construct the prompt
prompt_text = chatgpt_prompt(pname, search_links)
#get ChatGPT response
resp = get_chatgpt_resp(prompt_text)
# Create PDF with links and summary
rep_txt= generate_kyc_output(query, search_links, resp, start_time)
return (rep_txt)
main_tab, help_tab, rel_tab = st.tabs(["Run the Bot", "FAQ", "Release Plan"])
with main_tab:
# Streamlit app
st.title("Adverse News Detection Assistant")
# Input fields
names_txt = st.text_input("Enter party name (or multiple names separated by ,)")
plc_text = "laundering OR terrorist OR fraud OR corrupt OR criminal OR investigation OR prosecute OR evasion OR bribe OR sanction"
keywords = st.text_input("Enter other search words:", value=plc_text)
st.sidebar.markdown("## Controls")
st.sidebar.markdown("Choose your **search** *parameters*")
num_results = st.sidebar.slider("Choose the number of search results:", 5, 30, 20, 5)
st.sidebar.markdown("## Model")
st.sidebar.markdown("GPT v3.5")
st.sidebar.markdown("## App")
st.sidebar.markdown("v0.4")
col1, col2 = st.columns(2)
with col1:
adv_nw = st.radio(
"Did you find adverse news when you performed this search manually",
('Yes', 'No', 'Dont Know'), index=2)
with col2:
#st.markdown("Touch time (manual) in mins")
man_tt = st.number_input('Touch time (manual) in mins', value=0, step=1)
#st.markdown("Touch time (with bot) in mins")
bot_tt = st.number_input('Touch time (with bot) in mins', value=0, step=1)
# Search button
if st.button("Search"):
names = names_txt.split(",")
#print(len(names))
metrics_ent = (adv_nw != "Dont Know") and (man_tt > 0) and (bot_tt > 0)
# Perform the search and display the results
if names and metrics_ent:
search_results = ""
for name in names:
#print("trying for name {} \n".format(name))
search_results += perform_search(name, keywords, num_results)
html(f"<pre>{search_results}</pre>", height=200, scrolling=True)
st.download_button('Download Report',search_results)
try:
date_time = datetime.now()
save_to_s3(search_results,date_time )
print ("Completed processing for {} names: {} at {} \n".format(len(names), names_txt, str(date_time)))
except:
print ("Completed processing with S3 write error for {} names: {} at {} \n".format(len(names),names_txt, str(date_time)))
else:
st.error("Please enter party name, adverse news selection (Yes or No) and Touch Time before searching.")
with help_tab:
st.title("FAQ")
st.markdown("Q. How do I get a count of number of adverse news?")
st.markdown("A. This functionality isnt implemented yet. A workaround is to manually count the number of links with adverse news")
st.markdown("Q. How do I summarise all the adverse news?")
st.markdown("A. This functionality isnt implemented yet. A workaround is to aggregate the summary of all adverse news items manually, and get a sumary from ChatGPT (chat.openai.com")
st.markdown("Q. Can I search in other lauguages?")
st.markdown("A. This functionality isnt implemented yet. We are planning to test this feature out with Chinese first")
st.markdown("Q. Can I search without the other search words?")
st.markdown("A. Just enter a blank space in the text space and search")
with rel_tab:
st.markdown(f"""
| NO. | Issue / Enhancement | Rel | Status |
|-----|--------------------------------------------------------------------------------------------------------------------------------------------|-----|-----------|
| 1 | Capture productivity and adverse news metrics from the user | 0.4 | Completed |
| 2 | Save productivity and adverse news metrics in a DB | 0.4 | TBD |
| 3 | Convert bot output to structured JSON - Count of adverse news - Summary of all adverse news - Identification of links with adverse news | 0.6 | TBD |
| 4 | Offer alternate solution path with web text scraping and | 0.6 | TBD |
| 5 | Create a page on metrics report | 0.5 | TBD |""")
|