phd-proposal

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

PhD/Master's Research Proposal

博士/硕士研究计划书

Structure and write a rigorous research proposal that satisfies committee requirements.
构建并撰写符合委员会要求的严谨研究计划书。

Standard Proposal Structure (IEEE/ACM CS)

标准计划书结构(IEEE/ACM CS)

1. Title & Abstract          (~200 words)
2. Introduction              (~500 words)
3. Related Work              (~800 words)
4. Research Questions        (~200 words)
5. Methodology               (~1000 words) ← most important
6. Preliminary Results       (~500 words, if any)
7. Timeline & Milestones     (table/Gantt)
8. Expected Contributions    (~300 words)
9. References                (BibTeX)
1. Title & Abstract          (~200 words)
2. Introduction              (~500 words)
3. Related Work              (~800 words)
4. Research Questions        (~200 words)
5. Methodology               (~1000 words) ← most important
6. Preliminary Results       (~500 words, if any)
7. Timeline & Milestones     (table/Gantt)
8. Expected Contributions    (~300 words)
9. References                (BibTeX)

Section-by-Section Guidance

分章节指导

1. Title & Abstract

1. 标题与摘要

Title formula:
[Method/Approach] for [Problem] in [Domain/Context]
Example: "Federated Multi-Level Feedback Queue Scheduling for Resource-Constrained Healthcare IoT Networks"
Abstract (one paragraph, no citations, no math):
[Context]: In [domain], [background challenge].
[Gap]: However, [specific gap from literature].
[Approach]: We propose [method name], which [key idea].
[Validation]: We evaluate on [datasets/benchmarks] using [metrics].
[Impact]: Results show [expected outcome], enabling [application].
标题公式:
[方法/途径] 用于 [领域/场景] 中的 [问题]
示例: "面向资源受限医疗物联网网络的联邦多级反馈队列调度"
摘要(单段落,无引用,无数学公式):
[背景]: 在[领域]中,[背景挑战]。
[研究缺口]: 然而,[文献中的具体缺口]。
[研究方法]: 我们提出[方法名称],该方法[核心思路]。
[验证方式]: 我们在[数据集/基准测试集]上使用[指标]进行评估。
[影响]: 结果显示[预期成果],可实现[应用场景]。

2. Introduction

2. 引言

  • Hook: real-world motivation (1 paragraph)
  • Background: what exists (1 paragraph)
  • Problem: specific gap (1 paragraph)
  • Solution sketch: your approach (1 paragraph)
  • Contributions: bulleted list (use
    research-question
    output)
  • Thesis structure: "Chapter 2 covers..."
  • 切入点:现实场景动机(1段落)
  • 背景:现有研究情况(1段落)
  • 问题:具体研究缺口(1段落)
  • 方案概述:你的研究方法(1段落)
  • 研究贡献:项目符号列表(使用
    research-question
    技能输出内容)
  • 论文结构:“第2章涵盖……”

3. Related Work

3. 相关研究

Organize by taxonomy (not chronologically):
3.1 [Approach A] — summarize 3-5 papers, explain why insufficient
3.2 [Approach B] — same
3.3 [Your positioning] — "Unlike X, we address Y by Z"
按分类法组织(而非按时间顺序):
3.1 [方法A] — 总结3-5篇论文,说明其局限性
3.2 [方法B] — 同上
3.3 你的研究定位 — “与X不同,我们通过Z解决Y问题”

4. Research Questions

4. 研究问题

  • State main RQ + 2-4 sub-questions (paste from
    research-question
    skill output)
  • Each sub-question should be answerable in ~1 chapter/paper
  • 提出主研究问题(RQ)+2-4个子问题(粘贴
    research-question
    技能的输出内容)
  • 每个子问题应能在约1个章节/一篇论文中得到解答

5. Methodology (Most Critical)

5. 研究方法(最关键部分)

5.1 Research Paradigm Choose and justify:
  • Quantitative / Qualitative / Mixed
  • Paradigm: Design Science Research (DSR) / Experimental / Simulation
5.2 System/Model Design
Input: [what]
Processing: [how — algorithm, model, framework]
Output: [what]
Assumption: [what you assume]
5.3 Experimental Protocol
Datasets: [name, size, source, why chosen]
Baselines: [method1, method2 — why these are fair comparisons]
Metrics: [primary metric] + [secondary metrics]
Evaluation protocol: [k-fold CV / train-val-test split / simulation rounds]
Statistical tests: [t-test / Wilcoxon / effect size — for significance]
5.4 Reproducibility Plan
  • Code: GitHub repo (public after publication)
  • Environment: Docker / conda environment file
  • Seeds: fixed random seeds
  • Hardware: specify GPU/CPU specs
5.1 研究范式 选择并说明理由:
  • 定量研究 / 定性研究 / 混合研究
  • 范式:设计科学研究(DSR)/ 实验研究 / 模拟研究
5.2 系统/模型设计
输入:[内容]
处理过程:[方式——算法、模型、框架]
输出:[内容]
假设条件:[你的假设]
5.3 实验方案
数据集:[名称、规模、来源、选择理由]
基准方法:[方法1、方法2——选择这些作为公平对比的理由]
评估指标:[主要指标] + [次要指标]
评估方案:[k折交叉验证 / 训练-验证-测试集划分 / 模拟轮次]
统计检验:[t检验 / Wilcoxon检验 / 效应量——用于验证显著性]
5.4 可复现性计划
  • 代码:GitHub repo (public after publication)
  • 环境:Docker / conda environment file
  • 随机种子:固定随机种子
  • 硬件:明确GPU/CPU规格

6. Preliminary Results

6. 初步结果

Include if you have any early experiments. Even partial results strengthen the proposal.
  • 1-2 preliminary experiments showing feasibility
  • Small-scale validation of core idea
  • If none: state "Preliminary experiments are planned for Month 3-4"
若已有早期实验结果请附上。即使是部分结果也能提升计划书的说服力。
  • 1-2个证明可行性的初步实验
  • 核心思路的小规模验证
  • 若无结果:说明“初步实验计划在第3-4个月开展”

7. Timeline & Milestones

7. 时间线与里程碑

For a 2-year Master's program (align with
milestone-tracker
skill):
PhaseActivityDurationMilestone
1Literature reviewMonth 1-2Gap matrix complete
2Method designMonth 2-4Architecture finalized
3ImplementationMonth 3-8Working prototype
4ExperimentsMonth 7-14Results table complete
5WritingMonth 13-18Draft thesis/paper
6RevisionMonth 17-21Submission ready
7DefenseMonth 22-23Defended ✓
针对2年制硕士项目(与
milestone-tracker
技能对齐):
阶段活动时长里程碑
1文献综述第1-2个月完成研究缺口矩阵
2方法设计第2-4个月确定架构
3实现开发第3-8个月完成可运行原型
4实验开展第7-14个月完成结果表格
5论文撰写第13-18个月完成论文草稿
6修改完善第17-21个月准备好提交版本
7答辩第22-23个月答辩通过 ✓

8. Expected Contributions

8. 预期贡献

Restate from
research-question
skill, add:
  • Theoretical contribution (if any)
  • Practical contribution (system, dataset, tool)
  • Publication targets (venues + tier)
重述
research-question
技能的输出内容,并补充:
  • 理论贡献(如有)
  • 实践贡献(系统、数据集、工具)
  • 发表目标(会议/期刊 + 等级)

Committee Checklist

委员会评审 Checklist

Before submitting proposal for committee review:
  • Problem is clearly stated in 1-2 sentences
  • Gap is supported by ≥3 recent papers
  • RQ is specific and measurable
  • Methodology is detailed enough to replicate
  • Baselines are state-of-the-art, fairly chosen
  • Timeline is realistic (pad 20% for delays)
  • Limitations are acknowledged
  • Ethical considerations addressed (IRB, data privacy)
在提交计划书供委员会评审前:
  • 问题在1-2句话中清晰表述
  • 研究缺口由至少3篇近期论文支撑
  • 研究问题具体且可衡量
  • 研究方法详细到可复现
  • 基准方法为当前最优且选择公平
  • 时间线符合实际(预留20%时间应对延迟)
  • 已说明研究局限性
  • 已考虑伦理问题(IRB、数据隐私)

Output Format

输出格式

Produce a structured Markdown document following the section order above. Export to LaTeX using the
paper-writing
skill templates.
按照上述章节顺序生成结构化Markdown文档。使用
paper-writing
技能模板导出为LaTeX格式。

Links to Other Skills

与其他技能的关联

  • Requires →
    sota-survey
    +
    research-question
  • Feeds into →
    research-design
    (methodology details)
  • Feeds into →
    milestone-tracker
    (timeline)
  • Feeds into →
    paper-writing
    (proposal becomes thesis chapters)
  • 依赖 →
    sota-survey
    +
    research-question
  • 输出至 →
    research-design
    (研究方法细节)
  • 输出至 →
    milestone-tracker
    (时间线)
  • 输出至 →
    paper-writing
    (计划书成为论文章节)