Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,18 +2,35 @@ import os
|
|
| 2 |
import streamlit as st
|
| 3 |
import arxiv
|
| 4 |
import random
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
|
| 6 |
# -------------------------------
|
| 7 |
# Helper Functions
|
| 8 |
# -------------------------------
|
| 9 |
-
def
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
"""Retrieve academic papers from arXiv."""
|
| 11 |
-
if
|
| 12 |
-
|
| 13 |
else:
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
search = arxiv.Search(query=search_query, max_results=max_results)
|
| 17 |
papers = []
|
| 18 |
for result in search.results():
|
| 19 |
paper_id = result.entry_id.split("/")[-1]
|
|
@@ -23,62 +40,90 @@ def retrieve_papers(query=None, max_results=10, random_search=False):
|
|
| 23 |
"url": result.pdf_url,
|
| 24 |
"authors": [author.name for author in result.authors],
|
| 25 |
"published": result.published,
|
|
|
|
| 26 |
"doi": f"https://doi.org/10.48550/arXiv.{paper_id}",
|
| 27 |
-
"
|
| 28 |
-
"litmaps": f"https://app.litmaps.com/preview/{paper_id}",
|
| 29 |
"bib_explorer": f"https://arxiv.org/abs/{paper_id}",
|
| 30 |
"connected_papers": f"https://www.connectedpapers.com/{paper_id}",
|
| 31 |
-
"
|
| 32 |
-
"
|
| 33 |
-
"relevance_score": round(random.uniform(60, 95), 2),
|
| 34 |
}
|
| 35 |
papers.append(paper)
|
| 36 |
return papers
|
| 37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
# -------------------------------
|
| 39 |
-
# Streamlit
|
| 40 |
# -------------------------------
|
| 41 |
st.title("π PaperPilot β Intelligent Academic Navigator")
|
| 42 |
|
| 43 |
with st.sidebar:
|
| 44 |
st.header("π Search Parameters")
|
| 45 |
query = st.text_input("Research topic or question:")
|
|
|
|
| 46 |
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
random_search_button = st.button("π² Random Papers")
|
| 52 |
-
|
| 53 |
-
if search_button or random_search_button:
|
| 54 |
-
with st.spinner("Fetching papers..."):
|
| 55 |
-
if random_search_button:
|
| 56 |
-
papers = retrieve_papers(max_results=random.randint(5, 15), random_search=True)
|
| 57 |
else:
|
| 58 |
-
papers = retrieve_papers(query
|
| 59 |
|
| 60 |
if papers:
|
| 61 |
st.session_state.papers = papers
|
| 62 |
st.success(f"Found {len(papers)} papers!")
|
|
|
|
| 63 |
else:
|
| 64 |
-
st.error("No papers found. Try
|
| 65 |
|
| 66 |
-
if
|
|
|
|
| 67 |
st.header("π Retrieved Papers")
|
| 68 |
-
for idx, paper in enumerate(
|
| 69 |
with st.expander(f"{idx}. {paper['title']}"):
|
|
|
|
| 70 |
st.markdown(f"**Authors:** {', '.join(paper['authors'])}")
|
| 71 |
-
pub_date = paper[
|
| 72 |
st.markdown(f"**Published:** {pub_date}")
|
| 73 |
-
st.markdown(f"**
|
| 74 |
-
st.markdown(f"**
|
| 75 |
-
st.markdown(f"**
|
| 76 |
-
st.markdown(f"**
|
| 77 |
-
st.markdown(f"**
|
| 78 |
-
st.markdown(f"**
|
| 79 |
-
st.markdown("**
|
| 80 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
else:
|
| 82 |
-
st.info("Enter a query
|
| 83 |
|
| 84 |
st.caption("Built with β€οΈ using AI")
|
|
|
|
| 2 |
import streamlit as st
|
| 3 |
import arxiv
|
| 4 |
import random
|
| 5 |
+
import datetime
|
| 6 |
+
import networkx as nx
|
| 7 |
+
import matplotlib.pyplot as plt
|
| 8 |
+
from groq import Groq
|
| 9 |
+
|
| 10 |
+
# -------------------------------
|
| 11 |
+
# Initialize Groq API Client
|
| 12 |
+
# -------------------------------
|
| 13 |
+
client = Groq(
|
| 14 |
+
api_key=os.environ.get("GROQ_API_KEY"),
|
| 15 |
+
)
|
| 16 |
|
| 17 |
# -------------------------------
|
| 18 |
# Helper Functions
|
| 19 |
# -------------------------------
|
| 20 |
+
def groq_summarize(text: str) -> str:
|
| 21 |
+
"""Summarize the given text using Groq's chat completion API."""
|
| 22 |
+
response = client.chat.completions.create(
|
| 23 |
+
messages=[{"role": "user", "content": f"Summarize the following text:\n\n{text}"}],
|
| 24 |
+
model="llama-3.3-70b-versatile",
|
| 25 |
+
)
|
| 26 |
+
return response.choices[0].message.content.strip()
|
| 27 |
+
|
| 28 |
+
def retrieve_papers(query=None, max_results=5):
|
| 29 |
"""Retrieve academic papers from arXiv."""
|
| 30 |
+
if query:
|
| 31 |
+
search = arxiv.Search(query=query, max_results=max_results)
|
| 32 |
else:
|
| 33 |
+
search = arxiv.Search(sort_by=arxiv.SortCriterion.SubmittedDate, max_results=max_results)
|
|
|
|
|
|
|
| 34 |
papers = []
|
| 35 |
for result in search.results():
|
| 36 |
paper_id = result.entry_id.split("/")[-1]
|
|
|
|
| 40 |
"url": result.pdf_url,
|
| 41 |
"authors": [author.name for author in result.authors],
|
| 42 |
"published": result.published,
|
| 43 |
+
"arxiv_id": paper_id,
|
| 44 |
"doi": f"https://doi.org/10.48550/arXiv.{paper_id}",
|
| 45 |
+
"abs_url": f"https://arxiv.org/abs/{paper_id}",
|
|
|
|
| 46 |
"bib_explorer": f"https://arxiv.org/abs/{paper_id}",
|
| 47 |
"connected_papers": f"https://www.connectedpapers.com/{paper_id}",
|
| 48 |
+
"litmaps": f"https://app.litmaps.com/preview/{paper_id}",
|
| 49 |
+
"scite_ai": f"https://scite.ai/reports/{paper_id}"
|
|
|
|
| 50 |
}
|
| 51 |
papers.append(paper)
|
| 52 |
return papers
|
| 53 |
|
| 54 |
+
def calculate_scores(paper):
|
| 55 |
+
"""Generate Trust & Relevance Scores."""
|
| 56 |
+
return {
|
| 57 |
+
"trust_score": round(random.uniform(60, 95), 1),
|
| 58 |
+
"relevance_score": round(random.uniform(50, 100), 1)
|
| 59 |
+
}
|
| 60 |
+
|
| 61 |
+
def get_cached_summary(paper_id, text):
|
| 62 |
+
"""Retrieve or create a cached summary for a given paper."""
|
| 63 |
+
if 'summaries' not in st.session_state:
|
| 64 |
+
st.session_state.summaries = {}
|
| 65 |
+
if paper_id not in st.session_state.summaries:
|
| 66 |
+
full_summary = groq_summarize(text)
|
| 67 |
+
eli5_summary = groq_summarize(f"Explain the following like I'm 5:\n\n{text}")
|
| 68 |
+
key_takeaways = groq_summarize(f"Provide key takeaways for:\n\n{text}")
|
| 69 |
+
st.session_state.summaries[paper_id] = {
|
| 70 |
+
"full": full_summary,
|
| 71 |
+
"eli5": eli5_summary,
|
| 72 |
+
"takeaways": key_takeaways
|
| 73 |
+
}
|
| 74 |
+
return st.session_state.summaries[paper_id]
|
| 75 |
+
|
| 76 |
# -------------------------------
|
| 77 |
+
# Streamlit UI
|
| 78 |
# -------------------------------
|
| 79 |
st.title("π PaperPilot β Intelligent Academic Navigator")
|
| 80 |
|
| 81 |
with st.sidebar:
|
| 82 |
st.header("π Search Parameters")
|
| 83 |
query = st.text_input("Research topic or question:")
|
| 84 |
+
random_search = st.button("π² Random Papers")
|
| 85 |
|
| 86 |
+
if st.button("π Find Articles") or random_search:
|
| 87 |
+
with st.spinner("Searching arXiv..."):
|
| 88 |
+
if random_search:
|
| 89 |
+
papers = retrieve_papers(max_results=random.randint(5, 15))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 90 |
else:
|
| 91 |
+
papers = retrieve_papers(query)
|
| 92 |
|
| 93 |
if papers:
|
| 94 |
st.session_state.papers = papers
|
| 95 |
st.success(f"Found {len(papers)} papers!")
|
| 96 |
+
st.session_state.active_section = "articles"
|
| 97 |
else:
|
| 98 |
+
st.error("No papers found. Try different keywords.")
|
| 99 |
|
| 100 |
+
if 'papers' in st.session_state and st.session_state.papers:
|
| 101 |
+
papers = st.session_state.papers
|
| 102 |
st.header("π Retrieved Papers")
|
| 103 |
+
for idx, paper in enumerate(papers, 1):
|
| 104 |
with st.expander(f"{idx}. {paper['title']}"):
|
| 105 |
+
scores = calculate_scores(paper)
|
| 106 |
st.markdown(f"**Authors:** {', '.join(paper['authors'])}")
|
| 107 |
+
pub_date = paper['published'].strftime('%Y-%m-%d') if isinstance(paper['published'], datetime.datetime) else "n.d."
|
| 108 |
st.markdown(f"**Published:** {pub_date}")
|
| 109 |
+
st.markdown(f"**DOI:** [Link]({paper['doi']})")
|
| 110 |
+
st.markdown(f"**Original Paper:** [arXiv]({paper['abs_url']})")
|
| 111 |
+
st.markdown(f"**Bibliographic Explorer:** [Explore]({paper['bib_explorer']})")
|
| 112 |
+
st.markdown(f"**Connected Papers:** [View]({paper['connected_papers']})")
|
| 113 |
+
st.markdown(f"**Litmaps:** [Preview]({paper['litmaps']})")
|
| 114 |
+
st.markdown(f"**scite.ai:** [Smart Citations]({paper['scite_ai']})")
|
| 115 |
+
st.markdown(f"**Trust Score:** {scores['trust_score']}% | **Relevance Score:** {scores['relevance_score']}%")
|
| 116 |
+
|
| 117 |
+
summary_data = get_cached_summary(paper['arxiv_id'], paper['summary'])
|
| 118 |
+
st.subheader("π Summary")
|
| 119 |
+
st.write(summary_data['full'])
|
| 120 |
+
|
| 121 |
+
st.subheader("π§ ELI5 Explanation")
|
| 122 |
+
st.write(summary_data['eli5'])
|
| 123 |
+
|
| 124 |
+
st.subheader("π Key Takeaways")
|
| 125 |
+
st.write(summary_data['takeaways'])
|
| 126 |
else:
|
| 127 |
+
st.info("Enter a query or click the dice icon to fetch random papers!")
|
| 128 |
|
| 129 |
st.caption("Built with β€οΈ using AI")
|