Analysis API

Base path: /api/v1/projects/{project_id}/analysis

Endpoints

Method	Endpoint	Description
POST	/projects/{id}/analysis/contradictions	Detect contradictions across project papers
GET	/projects/{id}/analysis/author-network	Get author collaboration network
GET	/projects/{id}/analysis/trends	Get research trend analysis
POST	/projects/{id}/analysis/gaps	Detect literature gaps and research opportunities
GET	/projects/{id}/analysis/impact-scores	Get citation impact scores for papers

Contradictions

POST /projects/{id}/analysis/contradictions — Analyze all papers to find contradictory claims or findings using LLM.

Response fields:

• contradictions[] — Each with paper_a_id, paper_a_title, paper_b_id, paper_b_title, claim, position_a, position_b, confidence, topic
• topics[] — List of topics where contradictions were found
• total_contradictions — Count

Returns empty for projects with fewer than 2 papers.

Author Network

GET /projects/{id}/analysis/author-network — Build a co-authorship collaboration graph.

Query parameters:

• min_collaborations (default: 1) — Minimum co-authorship count for an edge
• max_nodes (default: 100, max: 500) — Maximum author nodes to return

Response fields:

• nodes[] — Each with name, paper_count, paper_ids, coauthors, h_index_estimate
• edges[] — Each with source, target, collaboration_count
• metrics — total_authors, total_edges, density, top_authors

Trends

GET /projects/{id}/analysis/trends — Analyze how research topics evolved over time.

Response fields:

• publication_timeline[] — Each with year, count, citations
• topic_trends[] — Each with topic, slope, r_squared, trend, total_papers, yearly_counts
• emerging_topics[] — Each with topic, yoy_growth
• declining_topics[] — Each with topic, yoy_growth
• summary_stats — total_papers, year_span, first_year, last_year, total_topics, emerging_count, declining_count

Gap Analysis

POST /projects/{id}/analysis/gaps — Identify under-researched areas and generate research questions using LLM.

Response fields:

• gaps[] — Each with topic, description, evidence, related_paper_ids, gap_score
• research_questions[] — Each with question, addresses_gap, novelty_score, feasibility_score
• summary — total_gaps, total_questions

Returns empty for projects with fewer than 2 papers.

Impact Scores

GET /projects/{id}/analysis/impact-scores — Compute Omelette Impact Score (0-100) for each paper.

Response fields:

• scores[] — Each with paper_id, title, score, factors (breakdown of score components)
• total — Number of scored papers
• avg_score — Average score across all papers
• top_paper_id — ID of the highest-scoring paper