Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -18,6 +18,9 @@ client = Groq(
|
|
| 18 |
# Helper Functions (Groq-based)
|
| 19 |
# -------------------------------
|
| 20 |
def groq_summarize(text: str) -> str:
|
|
|
|
|
|
|
|
|
|
| 21 |
response = client.chat.completions.create(
|
| 22 |
messages=[
|
| 23 |
{"role": "user", "content": f"Summarize the following text in detail:\n\n{text}"}
|
|
@@ -26,58 +29,67 @@ def groq_summarize(text: str) -> str:
|
|
| 26 |
)
|
| 27 |
return response.choices[0].message.content.strip()
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
"
|
| 46 |
-
|
|
|
|
| 47 |
|
| 48 |
-
# -------------------------------
|
| 49 |
-
# Retrieve Papers
|
| 50 |
-
# -------------------------------
|
| 51 |
def retrieve_papers(query, max_results=5):
|
|
|
|
| 52 |
search = arxiv.Search(query=query, max_results=max_results)
|
| 53 |
papers = []
|
| 54 |
for result in search.results():
|
| 55 |
-
paper_id = result.entry_id.split("/")[-1]
|
| 56 |
-
trust_score, relevance_score = calculate_trust_relevance(result)
|
| 57 |
-
external_links = get_external_links(paper_id)
|
| 58 |
paper = {
|
| 59 |
"title": result.title,
|
| 60 |
"summary": result.summary,
|
| 61 |
"url": result.pdf_url,
|
| 62 |
"authors": [author.name for author in result.authors],
|
| 63 |
"published": result.published,
|
| 64 |
-
"
|
| 65 |
-
"
|
| 66 |
-
"
|
|
|
|
|
|
|
| 67 |
}
|
| 68 |
papers.append(paper)
|
| 69 |
return papers
|
| 70 |
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
st.title("📚 PaperPilot – Intelligent Academic Navigator")
|
| 75 |
|
| 76 |
with st.sidebar:
|
| 77 |
st.header("🔍 Search Parameters")
|
| 78 |
query = st.text_input("Research topic or question:")
|
| 79 |
-
show_scores = st.checkbox("Show Trust & Relevance Scores", value=True)
|
| 80 |
-
|
| 81 |
if st.button("🚀 Find Articles"):
|
| 82 |
if query.strip():
|
| 83 |
with st.spinner("Searching arXiv..."):
|
|
@@ -85,30 +97,36 @@ with st.sidebar:
|
|
| 85 |
if papers:
|
| 86 |
st.session_state.papers = papers
|
| 87 |
st.success(f"Found {len(papers)} papers!")
|
| 88 |
-
st.session_state.active_section = "
|
| 89 |
else:
|
| 90 |
st.error("No papers found. Try different keywords.")
|
| 91 |
else:
|
| 92 |
st.warning("Please enter a search query")
|
| 93 |
|
|
|
|
|
|
|
|
|
|
| 94 |
if 'papers' in st.session_state and st.session_state.papers:
|
| 95 |
papers = st.session_state.papers
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
|
| 97 |
-
st.header("📑 Retrieved Papers")
|
| 98 |
-
for idx, paper in enumerate(papers, 1):
|
| 99 |
-
with st.expander(f"{idx}. {paper['title']}"):
|
| 100 |
-
st.markdown(f"**Authors:** {', '.join(paper['authors'])}")
|
| 101 |
-
pub_date = paper['published'].strftime('%Y-%m-%d') if isinstance(paper['published'], datetime.datetime) else "n.d."
|
| 102 |
-
st.markdown(f"**Published:** {pub_date}")
|
| 103 |
-
st.markdown(f"**Link:** [PDF Link]({paper['url']})")
|
| 104 |
-
if show_scores:
|
| 105 |
-
st.markdown(f"**Trust Score:** {paper['trust_score']:.2f} | **Relevance Score:** {paper['relevance_score']:.2f}")
|
| 106 |
-
st.markdown("**External Research Tools:**")
|
| 107 |
-
for tool, link in paper["external_links"].items():
|
| 108 |
-
st.markdown(f"- [{tool}]({link})")
|
| 109 |
-
st.markdown("**Abstract:**")
|
| 110 |
-
st.write(paper['summary'])
|
| 111 |
else:
|
| 112 |
st.info("Enter a query in the sidebar and click 'Find Articles' to get started.")
|
| 113 |
|
| 114 |
-
st.caption("Built with ❤️ using AI")
|
|
|
|
| 18 |
# Helper Functions (Groq-based)
|
| 19 |
# -------------------------------
|
| 20 |
def groq_summarize(text: str) -> str:
|
| 21 |
+
"""
|
| 22 |
+
Summarize the given text using Groq's chat completion API.
|
| 23 |
+
"""
|
| 24 |
response = client.chat.completions.create(
|
| 25 |
messages=[
|
| 26 |
{"role": "user", "content": f"Summarize the following text in detail:\n\n{text}"}
|
|
|
|
| 29 |
)
|
| 30 |
return response.choices[0].message.content.strip()
|
| 31 |
|
| 32 |
+
def groq_simplify(text: str) -> str:
|
| 33 |
+
"""
|
| 34 |
+
Explain Like I'm 5 (ELI5) version of the summary.
|
| 35 |
+
"""
|
| 36 |
+
response = client.chat.completions.create(
|
| 37 |
+
messages=[
|
| 38 |
+
{"role": "user", "content": f"Explain the following like I'm 5 years old:\n\n{text}"}
|
| 39 |
+
],
|
| 40 |
+
model="llama-3.3-70b-versatile",
|
| 41 |
+
)
|
| 42 |
+
return response.choices[0].message.content.strip()
|
| 43 |
|
| 44 |
+
def groq_generate_key_takeaways(text: str) -> str:
|
| 45 |
+
"""
|
| 46 |
+
Generate key takeaways from the paper.
|
| 47 |
+
"""
|
| 48 |
+
response = client.chat.completions.create(
|
| 49 |
+
messages=[
|
| 50 |
+
{"role": "user", "content": f"Provide key takeaways from this research paper:\n\n{text}"}
|
| 51 |
+
],
|
| 52 |
+
model="llama-3.3-70b-versatile",
|
| 53 |
+
)
|
| 54 |
+
return response.choices[0].message.content.strip()
|
| 55 |
|
|
|
|
|
|
|
|
|
|
| 56 |
def retrieve_papers(query, max_results=5):
|
| 57 |
+
"""Retrieve academic papers from arXiv, including DOI and tools for relevance/trust scoring."""
|
| 58 |
search = arxiv.Search(query=query, max_results=max_results)
|
| 59 |
papers = []
|
| 60 |
for result in search.results():
|
|
|
|
|
|
|
|
|
|
| 61 |
paper = {
|
| 62 |
"title": result.title,
|
| 63 |
"summary": result.summary,
|
| 64 |
"url": result.pdf_url,
|
| 65 |
"authors": [author.name for author in result.authors],
|
| 66 |
"published": result.published,
|
| 67 |
+
"doi": result.doi if hasattr(result, "doi") else f"https://doi.org/10.48550/arXiv.{result.entry_id.split('/')[-1]}",
|
| 68 |
+
"litmaps": f"https://app.litmaps.com/preview/{result.entry_id.split('/')[-1]}",
|
| 69 |
+
"connected_papers": f"https://www.connectedpapers.com/main/{result.entry_id.split('/')[-1]}",
|
| 70 |
+
"scite_ai": f"https://scite.ai/reports/{result.entry_id.split('/')[-1]}",
|
| 71 |
+
"biblio_explorer": f"https://arxiv.org/bib_explorer/{result.entry_id.split('/')[-1]}",
|
| 72 |
}
|
| 73 |
papers.append(paper)
|
| 74 |
return papers
|
| 75 |
|
| 76 |
+
def get_cached_summary(paper_id, text):
|
| 77 |
+
"""Retrieve or generate summaries, ELI5 explanations, and key takeaways."""
|
| 78 |
+
if 'summaries' not in st.session_state:
|
| 79 |
+
st.session_state.summaries = {}
|
| 80 |
+
if paper_id not in st.session_state.summaries:
|
| 81 |
+
st.session_state.summaries[paper_id] = {
|
| 82 |
+
"summary": groq_summarize(text),
|
| 83 |
+
"eli5": groq_simplify(text),
|
| 84 |
+
"key_takeaways": groq_generate_key_takeaways(text),
|
| 85 |
+
}
|
| 86 |
+
return st.session_state.summaries[paper_id]
|
| 87 |
+
|
| 88 |
st.title("📚 PaperPilot – Intelligent Academic Navigator")
|
| 89 |
|
| 90 |
with st.sidebar:
|
| 91 |
st.header("🔍 Search Parameters")
|
| 92 |
query = st.text_input("Research topic or question:")
|
|
|
|
|
|
|
| 93 |
if st.button("🚀 Find Articles"):
|
| 94 |
if query.strip():
|
| 95 |
with st.spinner("Searching arXiv..."):
|
|
|
|
| 97 |
if papers:
|
| 98 |
st.session_state.papers = papers
|
| 99 |
st.success(f"Found {len(papers)} papers!")
|
| 100 |
+
st.session_state.active_section = "review"
|
| 101 |
else:
|
| 102 |
st.error("No papers found. Try different keywords.")
|
| 103 |
else:
|
| 104 |
st.warning("Please enter a search query")
|
| 105 |
|
| 106 |
+
if 'active_section' not in st.session_state:
|
| 107 |
+
st.session_state.active_section = "none"
|
| 108 |
+
|
| 109 |
if 'papers' in st.session_state and st.session_state.papers:
|
| 110 |
papers = st.session_state.papers
|
| 111 |
+
|
| 112 |
+
if st.session_state.active_section == "review":
|
| 113 |
+
st.header("📚 Literature Review & Summary")
|
| 114 |
+
for idx, paper in enumerate(papers, 1):
|
| 115 |
+
with st.expander(f"Summary: {paper['title']}"):
|
| 116 |
+
with st.spinner(f"Analyzing {paper['title']}..."):
|
| 117 |
+
paper_id = f"paper_{idx}"
|
| 118 |
+
summary_data = get_cached_summary(paper_id, paper['summary'])
|
| 119 |
+
st.markdown(f"**Summary:** {summary_data['summary']}")
|
| 120 |
+
st.markdown(f"**ELI5:** {summary_data['eli5']}")
|
| 121 |
+
st.markdown("**Key Takeaways:**")
|
| 122 |
+
st.write(summary_data['key_takeaways'])
|
| 123 |
+
st.markdown(f"**DOI:** [Link]({paper['doi']})")
|
| 124 |
+
st.markdown(f"**Bibliographic Explorer:** [View]({paper['biblio_explorer']})")
|
| 125 |
+
st.markdown(f"**Connected Papers:** [View]({paper['connected_papers']})")
|
| 126 |
+
st.markdown(f"**Litmaps:** [View]({paper['litmaps']})")
|
| 127 |
+
st.markdown(f"**Scite.ai Citations:** [View]({paper['scite_ai']})")
|
| 128 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 129 |
else:
|
| 130 |
st.info("Enter a query in the sidebar and click 'Find Articles' to get started.")
|
| 131 |
|
| 132 |
+
st.caption("Built with ❤️ using AI")
|