researcher-writing-profiler

name

researcher-writing-profiler

description

Analyze a researcher's full publication record, writing style, and research strategy from their Google Scholar profile. Produces a structured Markdown report covering why they are prolific, their title/abstract/framing patterns, and actionable writing tips. Use this skill whenever the user asks to analyze a researcher's papers, understand why someone publishes so many top-venue papers, study an author's writing style, or wants to learn writing techniques from a prolific scholar. Triggers on: Google Scholar URLs, requests like "analyze this author's writing", "why does X publish so much at ACL/NeurIPS", "study this researcher's paper style", or any request to profile an academic's publication strategy.

Researcher Writing Profiler

Systematically analyze a researcher's entire publication record and writing style to extract actionable writing strategies. This skill turns a Google Scholar URL into a comprehensive "playbook" of how that researcher achieves prolific top-venue publications.

Input

The user provides one or more of:

• A Google Scholar profile URL (e.g., https://scholar.google.com/citations?user=XXXX)
• A researcher's name + affiliation
• Optionally, a target save location (e.g., an Obsidian vault folder)

Output

A structured Markdown report (2000-4000 words) saved to the user's specified location, covering:

1. Author profile & background
2. Publication statistics by venue and year
3. Research strategy analysis (why they are prolific)
4. Writing style deep analysis (title, abstract, framing, rhetorical techniques)
5. Actionable writing tips
6. Representative paper quick-reference table

---

Workflow

Phase 1: Data Collection (parallel where possible)

The goal is to gather two types of information: the full paper list and the author's background.

1A: Full Paper List from Google Scholar

Google Scholar is dynamically rendered and blocks simple HTTP fetching. Use the browser automation tools (Claude in Chrome / MCP) to navigate and scrape:

1. Get browser tab context (tabs_context_mcp), create a new tab if needed
2. Navigate to the Google Scholar profile URL
3. Use get_page_text to extract the initial paper list (first 20 papers)
4. Find and click the "Show more" button to load all papers — repeat until all papers are visible
5. Use get_page_text again to capture the full list

From the page, extract for each paper:

• Title
• Co-authors
• Venue (conference/journal name)
• Year
• Citation count

Also extract author-level metrics:

• Total citations (all time + since recent)
• h-index
• i10-index

If browser tools are unavailable, fall back to WebSearch + WebFetch to gather as much data as possible from Semantic Scholar API, DBLP, or ACL Anthology.

1B: Author Background

Search for the author's homepage (often linked from their Google Scholar profile) and fetch it:

1. WebSearch for "[Author Name] [Affiliation] homepage"
2. WebFetch the homepage to extract: education, advisors, research interests, awards, professional experience, co-author network

This background is crucial for understanding why they have certain research advantages (e.g., interdisciplinary training, strong advisor network).

1C: Representative Paper Abstracts

Fetch 4-6 abstracts from the author's most-cited or most representative papers across different venues. Use ArXiv, Semantic Scholar, or ACL Anthology. Focus on papers that received Oral/Spotlight designations or are from different research lines.

These abstracts are the raw material for writing style analysis.

Phase 2: Analysis

Analyze the collected data across five dimensions:

2A: Publication Statistics

Organize papers by:

• Venue tier: Count papers per top venue (ACL, EMNLP, ICLR, NeurIPS, ICML, CVPR, etc.)
• Year: Map annual output trajectory
• Venue type: Main conference vs. Findings vs. Workshop vs. arXiv

Present as tables for clarity.

2B: Research Strategy Analysis

Look for these patterns:

Pattern	What to look for
Interdisciplinary bridge	Papers combining two fields (e.g., psychology + NLP)
Benchmark/Evaluation route	Papers proposing evaluation frameworks rather than models
Research lines	Groups of 3-5 papers forming coherent series
Hot topic timing	Papers that align with emerging trends
Collaboration network	Distinct co-author clusters across institutions
Tool/Benchmark branding	Memorable names for frameworks (XxxBench, XxxAsker)

2C: Title Pattern Analysis

Classify every paper title into pattern categories. Common patterns include:

• Question titles: "Is X a Y?" / "X or Y?"
• Provocative statement + subtitle: "Bold claim: Technical description"
• Classical academic: "On the X of Y"
• Brand/Tool naming: Memorable coined names
• Cultural/Metaphor references: Adapted idioms or pop culture

Count frequency of each pattern and provide examples.

2D: Abstract Structure Analysis

Analyze 4-6 abstracts for their structural flow. Most strong abstracts follow a variant of:

1. Background/Context (1-2 sentences)
2. Gap/Problem identification (1-2 sentences)
3. Method/Contribution overview (2-3 sentences)
4. Key results with specific numbers (1-2 sentences)
5. Significance/Impact statement (1 sentence)

Note deviations and what makes this author's abstracts distinctive.

2E: Advanced Rhetorical Techniques

Across the papers, identify recurring techniques such as:

• Concrete-before-abstract: Leading with vivid examples
• Contrastive framing: Building arguments through contrast
• Quantitative anchoring: Using specific numbers for credibility
• Assumption-challenging framing: Questioning community consensus
• Zoom-in narrative: Progressive narrowing of scope in introductions
• Reveal arc: Not telegraphing conclusions in the introduction
• Spotlight analysis: Combining aggregate trends with case studies
• Two-level significance: Arguing importance at both practical and methodological levels
• Calibrated hedging: Strong claims where evidence supports, cautious elsewhere
• Preemptive objection handling: Addressing reviewer concerns proactively
• Inclusive scope signaling: Demonstrating comprehensive coverage

Phase 3: Report Generation

Compile everything into a structured Markdown report. Use the template below as a starting point, but adapt sections based on what's most distinctive about the specific researcher.

# [Author Name] 论文高产秘诀与写作风格分析

> 基于对其 Google Scholar 全部 [N] 篇论文的系统性分析
> 分析日期：[Date]

---

## 一、作者简介
[Table: name, affiliation, education, metrics, research areas]
[Key background insight that explains their research advantage]

## 二、发表记录统计
### 2.1 顶会分布 [Table]
### 2.2 年度产出 [Table]

## 三、为什么能中那么多顶会 —— 研究策略分析
### 3.1 [Most distinctive strategy]
### 3.2 [Second strategy]
### 3.3 [Third strategy]
...

## 四、写作风格深度分析
### 4.1 标题写作术 [Patterns with examples]
### 4.2 摘要写作术 [Structure analysis]
### 4.3 问题构框术 [Framing techniques]
### 4.4 举例与具象化
### 4.5 论文结构与叙事弧
### 4.6 贡献陈述术
### 4.7 高级修辞技巧 [Technique A, B, C...]

## 五、可实践的写作建议
[Numbered, actionable tips derived from the analysis]

## 六、代表作速查表
[Table: paper, venue, citations, core contribution, title technique]

## 七、总结
[Key takeaways]

Adapt language to match the user's preference (Chinese or English).

Phase 4: Save

Save the report to the user's specified location. If no location is specified, ask where they'd like it saved. Common targets:

• Obsidian vault folder
• Desktop
• A project directory

---

Key Principles

1. Be specific, not generic: Every claim should reference specific paper titles and concrete examples from the researcher's actual work. Avoid vague statements like "they write well."

2. Explain WHY, not just WHAT: Don't just catalog patterns — explain why each pattern works (e.g., "Question titles create curiosity gaps that increase click-through").

3. Actionable over descriptive: Every section should answer "how can I apply this to my own writing?"

4. Respect the data: If a researcher's success is partly due to non-replicable advantages (famous lab, massive compute), acknowledge this honestly alongside the replicable techniques.

5. Adapt to the researcher: Not all researchers are the same. A theorist's style will differ from an empiricist's. Adjust your analysis framework accordingly.

Notes

• Google Scholar may rate-limit or block requests. If the initial fetch fails, try loading more slowly or using alternative data sources.
• Some researchers have 100+ papers. Focus analysis depth on the top 20-30 most-cited papers while still cataloging the full list.
• If browser automation is unavailable, degrade gracefully: use WebSearch + WebFetch to gather what you can, and note any data gaps.
• The report language should match the user's language preference (auto-detect from conversation).