Update app.py

app.py

@@ -1,67 +1,82 @@
-# app.py
+# app.py – Now includes DuckDuckGo, arXiv, and Semantic Scholar crawling
 
 from sentence_transformers import SentenceTransformer
 from sklearn.metrics.pairwise import cosine_similarity
 import gradio as gr
+import arxiv
+from semanticscholar import SemanticScholar
+from duckduckgo_search import DDGS
 
-# Load fast transformer model
-model = SentenceTransformer('all-MiniLM-L6-v2')  # Faster & free
+# Load sentence transformer
+model = SentenceTransformer('all-MiniLM-L6-v2')
 
-# Define math domains and descriptions
+# Math domain definitions (trimmed for brevity)
 DOMAINS = {
     "Real Analysis": "Studies properties of real-valued functions, sequences, limits, continuity, differentiation, Riemann/ Lebesgue integration, and convergence in the real number system.",
-    "Complex Analysis": "Explores analytic functions of complex variables, contour integration, conformal mappings, and singularity theory.",
     "Functional Analysis": "Deals with infinite-dimensional vector spaces, Banach and Hilbert spaces, linear operators, duality, and spectral theory in the context of functional spaces.",
-    "Measure Theory": "Studies sigma-algebras, measures, measurable functions, and integrals, forming the foundation for modern probability and real analysis.",
-    "Fourier and Harmonic Analysis": "Analyzes functions via decompositions into sines, cosines, or general orthogonal bases, often involving Fourier series, Fourier transforms, and convolution techniques.",
-    "Calculus of Variations": "Optimizes functionals over infinite-dimensional spaces, leading to Euler-Lagrange equations and applications in physics and control theory.",
-    "Metric Geometry": "Explores geometric properties of metric spaces and the behavior of functions and sequences under various notions of distance.",
-    "Ordinary Differential Equations (ODEs)": "Involves differential equations with functions of a single variable, their qualitative behavior, existence, uniqueness, and methods of solving them.",
-    "Partial Differential Equations (PDEs)": "Deals with multivariable functions involving partial derivatives, including wave, heat, and Laplace equations.",
-    "Dynamical Systems": "Studies evolution of systems over time using discrete or continuous-time equations, stability theory, phase portraits, and attractors.",
-    "Linear Algebra": "Focuses on vector spaces, linear transformations, eigenvalues, diagonalization, and matrices.",
-    "Abstract Algebra": "General study of algebraic structures such as groups, rings, fields, and modules.",
-    "Group Theory": "Investigates algebraic structures with a single binary operation satisfying group axioms, including symmetry groups and applications.",
-    "Ring and Module Theory": "Extends group theory to rings (two operations) and modules (generalized vector spaces).",
-    "Field Theory": "Studies field extensions, algebraic and transcendental elements, and classical constructions.",
-    "Galois Theory": "Connects field theory and group theory to solve polynomial equations and understand solvability.",
-    "Algebraic Number Theory": "Applies tools from abstract algebra to study integers, Diophantine equations, and number fields.",
-    "Representation Theory": "Studies abstract algebraic structures by representing their elements as linear transformations of vector spaces.",
-    "Algebraic Geometry": "Examines solutions to polynomial equations using geometric and algebraic techniques like varieties, schemes, and morphisms.",
-    "Differential Geometry": "Studies geometric structures on smooth manifolds, curvature, geodesics, and applications in general relativity.",
-    "Topology": "Analyzes qualitative spatial properties preserved under continuous deformations, including homeomorphism, compactness, and connectedness.",
-    "Geometric Topology": "Explores topological manifolds and their classification, knot theory, and low-dimensional topology.",
-    "Symplectic Geometry": "Studies geometry arising from Hamiltonian systems and phase space, central to classical mechanics.",
-    "Combinatorics": "Covers enumeration, existence, construction, and optimization of discrete structures.",
     "Graph Theory": "Deals with the study of graphs, networks, trees, connectivity, and coloring problems.",
-    "Discrete Geometry": "Focuses on geometric objects and combinatorial properties in finite settings, such as polytopes and tilings.",
-    "Set Theory": "Studies sets, cardinality, ordinals, ZFC axioms, and independence results.",
-    "Mathematical Logic": "Includes propositional logic, predicate logic, proof theory, model theory, and recursion theory.",
-    "Category Theory": "Provides a high-level, structural framework to relate different mathematical systems through morphisms and objects.",
-    "Probability Theory": "Mathematical foundation for randomness, including random variables, distributions, expectation, and stochastic processes.",
-    "Mathematical Statistics": "Theory behind estimation, hypothesis testing, confidence intervals, and likelihood inference.",
-    "Stochastic Processes": "Studies processes that evolve with randomness over time, like Markov chains and Brownian motion.",
-    "Information Theory": "Analyzes data transmission, entropy, coding theory, and information content in probabilistic settings.",
-    "Numerical Analysis": "Designs and analyzes algorithms to approximate solutions of mathematical problems including root-finding, integration, and differential equations.",
+    "Partial Differential Equations (PDEs)": "Deals with multivariable functions involving partial derivatives, including wave, heat, and Laplace equations.",
     "Optimization": "Studies finding best outcomes under constraints, including convex optimization, linear programming, and integer programming.",
-    "Operations Research": "Applies optimization, simulation, and probabilistic modeling to decision-making problems in logistics, finance, and industry.",
-    "Control Theory": "Mathematically models and regulates dynamic systems through feedback and optimal control strategies.",
-    "Computational Mathematics": "Applies algorithmic and numerical techniques to solve mathematical problems on computers.",
-    "Game Theory": "Analyzes strategic interaction among rational agents using payoff matrices and equilibrium concepts.",
-    "Machine Learning Theory": "Explores the mathematical foundation of algorithms that learn from data, covering generalization, VC dimension, and convergence.",
-    "Spectral Theory": "Studies the spectrum (eigenvalues) of linear operators, primarily in Hilbert/Banach spaces, relevant to quantum mechanics and PDEs.",
-    "Operator Theory": "Focuses on properties of linear operators on function spaces and their classification.",
-    "Mathematical Physics": "Uses advanced mathematical tools to solve and model problems in physics, often involving differential geometry and functional analysis.",
-    "Financial Mathematics": "Applies stochastic calculus and optimization to problems in pricing, risk, and investment.",
-    "Mathematics Education": "Focuses on teaching methods, learning theories, and curriculum design in mathematics.",
-    "History of Mathematics": "Studies the historical development of mathematical concepts, theorems, and personalities.",
-    "Others / Multidisciplinary": "Covers problems that span multiple mathematical areas or do not fall neatly into a traditional domain."
+    "Algebraic Geometry": "Examines solutions to polynomial equations using geometric and algebraic techniques like varieties, schemes, and morphisms.",
+    "Others": "Covers problems that span multiple mathematical areas or do not fall neatly into a traditional domain."
 }
 
 domain_names = list(DOMAINS.keys())
 domain_texts = list(DOMAINS.values())
 domain_embeddings = model.encode(domain_texts)
 
+# Semantic Scholar + arXiv wrapper
+sch = SemanticScholar()
+
+def fetch_semantic_and_arxiv(query, top_n=5):
+    papers = []
+
+    # Semantic Scholar
+    try:
+        sem_results = sch.search_paper(query, limit=top_n)
+        for p in sem_results:
+            papers.append({
+                "title": p.title,
+                "authors": ", ".join(a.name for a in p.authors[:3]),
+                "year": p.year,
+                "url": p.url,
+                "source": "Semantic Scholar"
+            })
+    except:
+        pass
+
+    # arXiv
+    try:
+        search = arxiv.Search(query=query, max_results=top_n)
+        for r in search.results():
+            papers.append({
+                "title": r.title,
+                "authors": ", ".join(a.name for a in r.authors[:3]),
+                "year": r.published.year,
+                "url": r.entry_id,
+                "source": "arXiv"
+            })
+    except:
+        pass
+
+    return papers
+
+def fetch_duckduckgo_links(query, max_results=5):
+    links = []
+    try:
+        with DDGS() as ddgs:
+            results = ddgs.text(query, max_results=max_results)
+            for res in results:
+                links.append({
+                    "title": res['title'],
+                    "url": res['href'],
+                    "snippet": res['body'],
+                    "source": "DuckDuckGo"
+                })
+    except:
+        pass
+    return links
+
 def classify_math_question(question):
     q_embed = model.encode([question])
     scores = cosine_similarity(q_embed, domain_embeddings)[0]
@@ -70,16 +85,37 @@ def classify_math_question(question):
     minor = domain_names[sorted_indices[1]]
     major_reason = DOMAINS[major]
     minor_reason = DOMAINS[minor]
-    explanation = f"<b>Major Domain:</b> {major}<br><i>Reason:</i> {major_reason}<br><br>" + \
-                  f"<b>Minor Domain:</b> {minor}<br><i>Reason:</i> {minor_reason}"
-    return explanation
+
+    out = f"<b>Major Domain:</b> {major}<br><i>Reason:</i> {major_reason}<br><br>"
+    out += f"<b>Minor Domain:</b> {minor}<br><i>Reason:</i> {minor_reason}<br><br>"
+
+    refs = fetch_semantic_and_arxiv(question)
+    links = fetch_duckduckgo_links(question)
+
+    if refs:
+        out += "<b>Top Academic References:</b><ul>"
+        for p in refs:
+            out += f"<li><b>{p['title']}</b> ({p['year']}) - <i>{p['authors']}</i> [{p['source']}]<br><a href='{p['url']}' target='_blank'>{p['url']}</a></li>"
+        out += "</ul>"
+    else:
+        out += "<i>No academic references found.</i><br>"
+
+    if links:
+        out += "<b>Top Web Resources:</b><ul>"
+        for link in links:
+            out += f"<li><b>{link['title']}</b><br>{link['snippet']}<br><a href='{link['url']}' target='_blank'>{link['url']}</a></li>"
+        out += "</ul>"
+    else:
+        out += "<i>No web links found.</i>"
+
+    return out
 
 iface = gr.Interface(
     fn=classify_math_question,
-    inputs=gr.Textbox(lines=5, label="Enter Math Question (supports LaTeX)"),
-    outputs=gr.HTML(label="Predicted Domains"),
-    title="🔍 Math Domain Classifier (Major + Minor)",
-    description="Paste any math problem or LaTeX expression and get its major and minor domains with explanations."
+    inputs=gr.Textbox(lines=5, label="Enter Math Question (LaTeX supported)"),
+    outputs=gr.HTML(label="Predicted Domains + References"),
+    title="📚 Math Domain Classifier with arXiv + Semantic Scholar + DuckDuckGo",
+    description="Paste any math problem or LaTeX expression and get its major/minor domains + academic references + web links."
 )
 
 iface.launch()