Upload 4 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+papers_index.faiss filter=lfs diff=lfs merge=lfs -text

Dockerfile

@@ -0,0 +1,8 @@
+FROM python:3.11
+
+WORKDIR /app
+COPY . /app
+RUN pip install --no-cache-dir -r requirements.txt
+
+EXPOSE 7860
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,115 @@
+from fastapi import FastAPI, HTTPException, Query
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+import pandas as pd
+import numpy as np
+import faiss
+from sentence_transformers import SentenceTransformer
+import os
+import gdown
+
+app = FastAPI(title="Research Paper Recommendation API")
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],  # Allow Next.js frontend
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+print("🔄 Loading dataset and FAISS index...")
+def download_from_drive(file_id, output):
+    if not os.path.exists(output):
+        print(f"📥 Downloading {output} from Google Drive...")
+        url = f"https://drive.google.com/uc?id={file_id}"
+        gdown.download(url, output, quiet=False)
+    else:
+        print(f"✅ {output} already exists, skipping download.")
+
+download_from_drive("1ME6Bb5WjVbIYr4-0iF0kUa35-82DTsUn", "papers_with_embeddings.csv")
+download_from_drive("1IwOkhBu-odM2GYmvZdS1Q5AHkf0pDV02", "embeddings.npy")
+download_from_drive("1MEB_4ZGunbxi65jW9UN0L8bnp9VUCH8s", "papers_index.faiss")
+
+try:
+    df = pd.read_csv("papers_with_embeddings.csv")
+    embeddings = np.load("embeddings.npy")
+    index = faiss.read_index("papers_index.faiss")
+except Exception as e:
+    raise RuntimeError(f" Failed to load files: {e}")
+
+model = SentenceTransformer("allenai/specter2_base")
+
+print(f" Loaded {len(df)} papers and FAISS index with {index.ntotal} vectors.")
+
+class Paper(BaseModel):
+    id: str
+    title: str
+    authors: str
+    update_date: str
+    abstract: str | None = None
+    category_code: str | None = None
+
+def get_recommendations(paper_id: str, top_k: int = 6):
+    paper = df[df["id"].astype(str) == str(paper_id)]
+    if paper.empty:
+        raise HTTPException(status_code=404, detail="Paper not found")
+
+    text = f"{paper.iloc[0]['title']}. {paper.iloc[0]['abstract']}"
+    query_vec = model.encode([text], normalize_embeddings=True)
+
+    D, I = index.search(query_vec, top_k + 1)
+    recs = df.iloc[I[0]].copy()
+    recs["similarity"] = D[0]
+    # Exclude the query paper itself
+    recs = recs[recs["id"].astype(str) != str(paper_id)]
+
+    return recs[["id", "title", "authors", "update_date", "abstract", "similarity"]].head(top_k).to_dict(orient="records")
+
+
+def search_papers(query_text: str, top_k: int = 50):
+    if not query_text.strip():
+        raise HTTPException(status_code=400, detail="Query cannot be empty.")
+
+    query_vec = model.encode([query_text], normalize_embeddings=True)
+    D, I = index.search(query_vec, top_k)
+
+    recs = df.iloc[I[0]].copy()
+    recs["similarity"] = D[0]
+    # Include abstract in output
+    return recs[["id", "title", "authors", "update_date", "abstract", "similarity"]].to_dict(orient="records")
+
+
+@app.get("/")
+def root():
+    return {"message": "SPECTER + FAISS Recommendation API is running 🚀"}
+
+@app.get("/paper/{paper_id}")
+def get_paper(paper_id: str):
+    paper = df[df["id"].astype(str) == str(paper_id)]
+    if paper.empty:
+        raise HTTPException(status_code=404, detail="Paper not found")
+    return paper.iloc[0].to_dict()
+
+@app.get("/recommend/{paper_id}")
+def recommend_papers(paper_id: str, top_k: int = 6):
+    try:
+        recs = get_recommendations(paper_id, top_k)
+        return recs
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Recommendation error: {e}")
+
+@app.get("/search")
+def search_endpoint(query: str = Query(..., description="Search text query"), top_k: int = 50):
+    """
+    Search for semantically similar papers using SPECTER embeddings.
+    Example: /search?query=graph neural networks
+    """
+    try:
+        results = search_papers(query, top_k)
+        return results
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Search error: {e}")
+

papers_index.faiss

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fb48220b52b84c794b9409e70e4df599500f6a11f8d6e68c67630cfdb125455
+size 596167725

requirements.txt

@@ -0,0 +1,7 @@
+fastapi
+uvicorn
+pandas
+numpy
+faiss-cpu
+sentence-transformers
+gdown