| name | ai4scholar-paper-recommendation |
| description | Use AI4Scholar recommendation tools to find related papers from one or more seed papers, build reading lists, expand a literature map, and discover papers adjacent to a user's theory, method, dataset, or empirical context. |
AI4Scholar Paper Recommendation
Use this skill when the user has seed papers and wants targeted related-paper recommendations rather than a broad keyword search.
Core Rules
- Recommendations are candidates, not verified citations.
- Always keep the seed set visible so the recommendation logic can be audited.
- Do not let recommendations drift away from the user's research question.
- Deduplicate recommendations against papers already in the user's library.
Tools From The Source Guide
| Tool | Use |
|---|
get_semantic_recommendations | Recommend papers based on multiple seed papers |
get_semantic_recommendations_for_paper | Recommend papers based on a single seed paper |
get_semantic_paper_detail | Verify details for recommended papers |
get_semantic_citations / get_semantic_references | Combine recommendation with citation network expansion |
Recommendation Workflow
- Ask for or infer the seed papers:
- DOI,
- Semantic Scholar ID,
- arXiv ID,
- exact title.
- Classify each seed by role:
- theory,
- method,
- empirical context,
- dataset,
- measurement,
- modeling technique.
- Run recommendation tools.
- Deduplicate with current reading list.
- Screen recommendations for:
- topic fit,
- method fit,
- venue and year,
- whether it adds something new beyond the seed papers.
- Return a reading order.
Use Cases
- "I have 5 papers, what should I read next?"
- "Find papers similar to this ERP/EEG experiment."
- "Find BCI deep learning papers near this architecture."
- "Find management/psychology papers using similar scenario experiments."
- "Expand a literature review without changing the research question."
Output Template
Seed papers:
Recommendation tool:
Recommended papers:
| Priority | Title | Authors | Year | Venue | DOI/ID | Why recommended | What to read for |
Not selected:
| Title | Reason |
Reading order:
Beginner Prompt
Use AI4Scholar to recommend follow-up readings from the following seed papers.
Do not rely only on keywords; explain the recommendation by theory, method, data, and research context.
Return priority, DOI/ID, reading purpose, and verification status.
Quality Checks
- If the seed paper is too broad, ask for the user's target angle.
- If recommendations are mostly irrelevant, switch to
ai4scholar-paper-search with narrower queries.
- If the user needs a complete literature map, combine this skill with
ai4scholar-citation-network.