# 课题组逐院系 · 完备性重做 · 贝尔法斯特女王大学（Queen's University Belfast）· `qub.ac.uk`

- 日期：2026-09-23 · Manager 波次 · 单校 · **只读**
- 方式：ego-browser（**自建 TaskSpace 75**，每个 URL 用独占的 `newPage()`，逐页校验 `page.url()` 与目标一致；成功时收尾 `await task.finish({ keep: [] })`）+ Node.js fetch 并行爬取做规模（限速 2 并发 + 0.9s 间隔）
- 前一轮：`finding_qub.md`（29 组 / 16 个 School = 1.8）——**本报告推翻其结论**：上一轮只沿 `/research-centres/` 校级总表收「中心」，**完全没进 15 个 School 各自的 research 子树**，也没收中心内部具名组、`/sites/` PI 组站、NBE Research Cells。
- 产物：
  - `qub_v2__groups.json`（**133 条** 五字段）
  - `qub_v2__rows_all.json`（204 条候选 + 每条 `included`/`exclude_reason`）
  - `qub_v2__ledger.md`（§1 完备性台账）
  - 本文

---

## 0. 结论先行

**这所学校最终组数是 133（其中 117 条归属官方 15 个 School，另 16 条是研究所与校级中心的具名内层单元）。**

前一轮 29 组**不是判据问题，是覆盖问题**：它只收了校级研究中心总表一层，漏掉了
School 自己的 research 子树（15 个 School 的 research 导航）、
中心内部的具名组（MDBS 三中心下的 10 个组 + 3 个 Centre of Excellence、QUILL 的 10 个 PI 组、CPH 的 4 个 Research Group）、
以及 `/sites/` 下的 PI 组站（ACRG、Science and Culture Research Group 等）。

---

## 1. 完备性台账（官方 **15 个 School** + 研究所/校级中心）

**枚举来源 URL**：
- `https://www.qub.ac.uk/schools/`（官方 University Schools：Schools by Faculty 15 条 + Other Academic Provision A–Z 8 条）
- `https://www.qub.ac.uk/about/Leadership-and-structure/Faculties-and-Schools/`（原文 “Queen's three Faculties and **fifteen Schools**”）
- `https://www.qub.ac.uk/about/Leadership-and-structure/University-department-list/`（院系/中心/研究所清单）
- `https://www.qub.ac.uk/research-centres/`（University Research Centres 总表，列 ~80 条）
- 各 School 的 research 导航（下表第 3 列）+ GRI 页 `https://www.qub.ac.uk/Research/GRI/`

| # | 院系/单位 | research 入口 URL | 状态 | 组数 | 备注 |
|---|---|---|---|---|---|
| 1 | School of Arts, English and Languages | `https://www.qub.ac.uk/schools/ael/Research/` | 取到 | 8 | Research in Arts / English / Languages 三个子索引 + `/sites/ScienceandCultureResearchGroup/` |
| 2 | School of History, Anthropology, Philosophy and Politics | `https://www.qub.ac.uk/schools/happ/subject-area/*/research/` | 取到 | 8 | 该 School **无** `/schools/happ/Research/`（404）；按 4 个 subject-area 的 research 页枚举 |
| 3 | School of Law | `https://www.qub.ac.uk/schools/SchoolofLaw/research/` | 取到 | 10 | 6 个 thematic area + 4 个中心/研究所 |
| 4 | Queen's Business School | `https://www.qub.ac.uk/schools/queens-business-school/research/` | 取到 | 4 | `research-centres-and-initiatives` 页 |
| 5 | School of Social Sciences, Education and Social Work | `https://www.qub.ac.uk/schools/ssesw/research/` | 取到 | 8 | 8 个 centre |
| 6 | School of Chemistry and Chemical Engineering | `.../SchoolofChemistryandChemicalEngineering/Research/ResearchCentres/` | 取到 | 4 | CASE / CO2@QUB / QUILL / PPRC |
| 7 | School of Electronics, Electrical Engineering and Computer Science | `https://www.qub.ac.uk/schools/eeecs/Research/` | 取到 | 5 | `eeecs-research-centres` 页 5 个中心（CISC/CWI/EPIC/CSIT/CES） |
| 8 | School of Mathematics and Physics | `.../SchoolofMathematicsandPhysics/Research/ResearchCentres/` | 取到 | 4 | 页称 “six research centres”；逐页取到 4 个有独立页的中心（ARC/CQMT/CLMI/MSRC） |
| 9 | School of Mechanical and Aerospace Engineering | `.../SchoolofMechanicalandAerospaceEngineering/Research/` | 取到 | 7 | 6 theme + ACRG `/sites/` 组站 |
| 10 | School of Natural and Built Environment | `https://www.qub.ac.uk/schools/NBE/Research/` | 取到 | 13 | ResearchCells（8 个 PI-cell）+ research-centres-and-institutes |
| 11 | School of Psychology | `https://www.qub.ac.uk/schools/psy/Research/` | 取到 | 9 | 3 theme 下的 lab/centre |
| 12 | School of Biological Sciences | `.../SchoolofBiologicalSciences/Research/` | 取到 | 1 | 3 theme（大一级）+ 1 Centre of Excellence；theme 是院里最细一层 |
| 13 | School of Medicine, Dentistry and Biomedical Sciences | `https://www.qub.ac.uk/schools/mdbs/Research/` | 取到 | 18 | 3 centre（PGJCCR/WWIEM/CPH）内层组 10 + 3 Centre of Excellence + 3 院内 centre |
| 14 | School of Nursing and Midwifery | `.../SchoolofNursingandMidwifery/Research/research-themes/` | 取到 | 3 | 3 research theme（各有 Theme Lead 与团队） |
| 15 | School of Pharmacy | `.../SchoolofPharmacy/Research/ResearchThemes/` | 取到 | 15 | 5 theme × 共 15 个 area of activity（每 area 列具名 PI） |
| 16 | Institute for Global Peace, Security and Justice（GRI） | `.../institute-for-global-peace-security-justice/ResearchandImpact/ResearchPriorities/` | 取到 | 8 | 8 个 research priority（institute 内层具名单元） |
| 17 | Institute for Global Food Security（GRI） | `https://www.qub.ac.uk/Research/GRI/TheInstituteforGlobalFoodSecurity/research/` | 取到 | 4 | 4 个 research theme |
| 18 | University Research Centres（校级总表） | `https://www.qub.ac.uk/research-centres/` | 取到 | 4 | 总表 ~80 条；校级/登记册/项目/设施按判据剔除，逐页取到 4 条真组 |
| | **合计** | | | **133** | 15 School 归属 **117** + 非 School 单位 16 |

**未取页 = 0**（15 个 School 的 research 入口全部 200 取到）。取不到但有证据的逐条：

| 目标 | 结果 | 错误原文/证据 |
|---|---|---|
| `/schools/happ/Research/` | 404 | 页面 H1 = “PAGE NOT FOUND”；该校 HAPP 改由 subject-area 页进入 |
| `http://www.am.qub.ac.uk/ctamop/` | 重定向 | 最终 URL 回到 `/schools/SchoolofMathematicsandPhysics/Research/ResearchCentres/`（CTAMOP 已无独立页） |
| `http://www.am.qub.ac.uk/` | jump 页 | “Redirecting Please wait a few seconds…” |
| `https://pure.qub.ac.uk/en/organisations/` | 403 | Cloudflare “Just a moment…” JS 挑战（未用 Pure 做枚举） |

---

## 2. 子域 / 外站清单（**已跟**）

本次抓到 **39 个 `*.qub.ac.uk` 子域**。承载研究单元的：

`www.qub.ac.uk`（主站）· `www.am.qub.ac.uk` / `web.am.qub.ac.uk`（物理，重定向）· `www.sarc.qub.ac.uk` · `www.ee.qub.ac.uk` · `www.med.qub.ac.uk` · `www.erin.qub.ac.uk` · `www.socasites.qub.ac.uk` · `blogs.qub.ac.uk` · `qpol.qub.ac.uk` · `law.qub.ac.uk` · `edgehub.eeecs.qub.ac.uk` · `irishsongproject.qub.ac.uk` · `reparations.qub.ac.uk` · `nicola.qub.ac.uk` · `cmst4.qub.ac.uk` · `coe.qub.ac.uk` · `dap.qub.ac.uk` · `star.pst.qub.ac.uk` · `quote.qub.ac.uk` · `omeka.qub.ac.uk` · `pure.qub.ac.uk`（Pure 门户，403 未跟）。

**非研究单位**（列出以便核对）：`alumni / myportal / vle.qol / mediasite / qmail / hr.apps / hrwebapp / ecommerce.apps / go / daro / intranet.qol / virtualexperience / libguides / libcal / libraryhelp`（图书馆与行政域）。

**承载研究单元的外站**（已跟）：
`ark.ac.uk`（ARK 社会政策中心）· `ceph.ie`（CEPH）· `riscplus.com`（RiSC+）· `www.productivity.ac.uk`（NI Productivity Forum）· `14chrono.org`（Centre for Climate, the Environment & Chronology，实测为**定年服务实验室**，判为设施不收）· `www.leadershipinstitute.co.uk`（执行教育）· `www.prisonsmemoryarchive.com` · `recomposingthecity.org` · `translatingimprovisation.com` · `culturalpolicyireland.org` · `pcancities.org.uk` · `www.detect-project.eu`。

> 说明：整站抓取共引用 **1673 个外站主机**，其中绝大多数是论文/出版社/会议/新闻引用（`academic.oup.com`、`arxiv.org`、`gtr.ukri.org`、`apnews.com`、`wiley.com` …），**不是研究单位**，不逐条跟。上面的清单是当中真正承载 QUB 研究单元的站点。

---

## 3. 组 / 系 比值与合理性校验

- 15 个 School，逐页核到 **133** 组（15 School 归属 117，非 School 单位 16）。
- **组/系 ≈ 8.9**（仅 School 归属 ≈ 7.8）。基准：布里斯托 ≈5.8、曼彻斯特 ≈8.4。**不荒谬地低，无需再往下挖一层。**
- QUB 的院系确实大量用 theme/centre 组织（大一级），但**底层具名单元本次已收**：

| 院系 | 大一级组织 | 底下挂着的具名组（本次已收） |
|---|---|---|
| MDBS | 3 个 Research Centre（PGJCCR / WWIEM / CPH） | CPH 4 个 Research Group；PGJCCR 6 个癌种组 + 3 个 Centre of Excellence；WWIEM 3 个 theme（含 “16 PI-led teams”） |
| Chemistry | 3 个 Research Centre | **QUILL 10 个 PI 组**（页面原文 “brings together **ten research groups led by independent principal investigators**”） |
| NBE | 5 个 Discipline + 3 个 Cluster | **8 个 Research Cell**（原文 “Each Research Cell is headed by a Professor or Senior Academic”） |
| Psychology | 3 个 Theme | 7 个 lab/centre（CIHRQoL / STARC / CIRR / Kids in Context / Animal Behaviour / InteRRaCt / Social Interactions Lab） |
| Pharmacy | 5 个 Theme | **15 个 area of activity**（每 area 列具名 PI） |
| IGPSJ（GRI） | 1 个 Institute | **8 个 research priority** |
| IGFS（GRI） | 1 个 Institute | 4 个 research theme |
| M&P | 4 个 Research Centre | ARC 6 / CLMI 9 / CQMT 3 / MSRC 8 个内层 sub-theme（**均无独立页、无逐单元方向文字，故不收**，只计 4 个中心） |
| EEECS | 5 个 Centre | CWI 6 sub-theme、CSIT 3 cluster（**同上，只计 5 个中心**） |

- **“再下一层没有组”的原文结构佐证**：
  - School of Chemistry `/Research/ResearchCentres/` 只有三个中心（CASE / CO2@QUB / QUILL），QUILL 的 PI 组已计入，无更细一层。
  - M&P `/Research/ResearchCentres/` 原文 “Our **six** research centres”，逐页只取到 4 个有独立页的；另 2 个（CTAMOP、Centre for Plasma Physics）无独立页（见 §5）。
  - School of Biological Sciences `/Research/ResearchThemes/`：H1 = “Research Themes”，正文只有 3 个 theme + Theme Leads，**没有下钻的组页**（原文 “Our Three Research Themes: …”）。

---

## 4. 启发式自动抽 + 抽样 30 条

### 4.1 启发式自动抽（全量，204 条候选）

机械判据：页面 URL/H1 指向具名研究单元（centre / group / lab / cluster / cell / strand / theme）**且**页面里有**方向文字块**（>60 字，去掉导航与 “Link -” 后仍成立）**且**名字不是索引页/课程/项目/学生群体/讲座系列/设施。自述含 `N research groups led by independent PIs` / `N-strong team` / institute 级 ⇒ 标 `big`，只收其内层具名单元。
→ 产出候选 **204 条**，写入 `qub_v2__rows_all.json`。

### 4.2 第一轮抽样 30 条（原始候选池 204，逐条开页面核）

| # | 类型 | 名称 | 目标 URL | 判定 | 依据（打开页面看到什么） |
|---|---|---|---|---|---|
| 1 | group | Breast and Ovarian Cancer | `/research-centres/cancer-research/OurResearch/BreastCancerOvarianCancer/` | ✔ | 独立页 H1 “Breast Cancer & Ovarian Cancer”，有方向 |
| 2 | group | Proteases in Disease | `/schools/SchoolofPharmacy/Research/ResearchThemes/` | ✘ | 共享索引页列表项，目标页无 “Proteases in Disease” 方向文字 |
| 3 | theme | Advanced Manufacturing and Robotics | `.../Research/advanced-manufacturing/` | ✔ | 独立 theme 页，有 mission 与方向 |
| 4 | theme | Connected & Autonomous Systems | `/research-centres/cwi/about/` | ✘ | CWI 索引页上的 6 个 sub-theme 之一，仅列表名，无本单元方向 |
| 5 | centre | ARK（Access Research Knowledge） | `https://www.ark.ac.uk/` | ✘ | 社会调查数据/资源 hub，非导师带的组 |
| 6 | centre | Northern Ireland Productivity Forum | `productivity.ac.uk/...` | ✔ | 独立页，有方向 |
| 7 | group | Epidemiology and Public Health Research Group | `.../CentreforPublicHealth/Research/EpidemiologyPublicHealth/` | ✔ | 独立组页，有 lead 与成员 |
| 8 | theme | Green Wireless | `/research-centres/cwi/about/` | ✘ | 同 #4：CWI sub-theme 列表项 |
| 9 | centre | Centre for Eighteenth-Century Studies | `/schools/CentreforEighteenth-CenturyStudies/` | ✔ | 独立页，interdisciplinary research initiative |
| 10 | group | Advanced Composites Research Group | `/sites/acrg/` | ✔ | 独立组站，原文 “led by Dr Zafer Kazancı” |
| 11 | centre | Northern Ireland Technology Centre | `/sites/nitc/` | ✘ | 技术/创新**服务中心**（design, knowledge engineering, manufacturing services） |
| 12 | cell | Futures Presents Pasts | `/schools/NBE/Research/ResearchCells/` | ✔ | NBE Research Cell，共享页有逐 cell 方向文字与 cell lead |
| 13 | centre | QUILL Research Centre | `.../Research/QUILL/` | ✔ | 独立中心页，原文 “ten research groups led by independent PIs” |
| 14 | lab | InteRRaCt Lab | `/schools/psy/Research/ResearchFacilities/` | ✘ | URL **404**（该页已失效） |
| 15 | centre | Centre for Language Education Research | `/research-centres/CentreforLanguageEducationResearch/` | ✔ | 独立中心页 CLER |
| 16 | centre | Centre for European and Transnational Studies | `/research-centres/centre-european-transnational-studies/` | ✔ | 独立中心页 |
| 17 | theme | Sustainable Energy, Agriculture and Transport Systems | `.../Research/energy-agriculture-transport/` | ✔ | 独立 theme 页，有 mission |
| 18 | institute | Institute of Cognition and Culture | `/schools/InstituteofCognitionCulture/` | ✘ | **institute 粒度（大一级）**；其下只有 project，按判据不收 |
| 19 | theme | Particle acceleration and secondary sources | `/research-centres/light-matter-interactions/Researchthemes/` | ✘ | 共享索引页列表项，无本单元方向 |
| 20 | centre | MethodsLab | `/research-centres/methodslab/` | ✘ | 原文 “methodological **support** and innovation resource”，服务/支持资源 |
| 21 | centre | CEAM | `.../CentreofExcellenceforAgri-FoodMicrobiomesCEAM/` | ✔ | 独立页 |
| 22 | centre | QUOTE | `https://quote.qub.ac.uk/` | ✔ | 有研究方向自述 |
| 23 | centre | Human Rights Centre | `/research-centres/human-rights-centre/` | ✔ | 独立中心页 |
| 24 | group | Science and Culture Research Group | `/sites/ScienceandCultureResearchGroup/` | ✔ | 独立组站，有方向与 3 位 lead |
| 25 | group | Respiratory Infectious Diseases | `/schools/SchoolofPharmacy/Research/ResearchThemes/` | ✘ | URL 指向 theme 索引页，该页 **无**此 area 的方向文字（**修 URL 后转 ✔**） |
| 26 | centre | Centre for Dentistry | `/schools/mdbs/centre-for-dentistry/` | ✘ | 教学/临床中心，原文 “where your future in dentistry begins” |
| 27 | cluster | Private and Commercial Law | `/schools/SchoolofLaw/research/private-commercial-law/` | ✔ | Law 6 个 thematic area 之一，有方向与 affiliated staff |
| 28 | lab | Social Interactions Lab | `/schools/psy/.../SocialPsychologyandtheStudyofSocialIssues/` | ✔ | Psy theme 页列出的 hub |
| 29 | strand | Quantum Technologies | `/research-centres/quantum-materials-technologies/research-strands/` | ✘ | CQMT strands 索引页列表项，无本 strand 方向 |
| 30 | centre | Reimagining Supply Chains Network Plus | `https://riscplus.com/` | ⚠ | 触发 captcha，**未取到** ⇒ 未验 |

**第一轮准确率（剔除未验 1 条）：17 / 29 ≈ 58.6%**（✔ 17，✘ 12，⚠ 1）。

**假阳性类型（集中，不是散的）：**

| 类型 | 条数 | 页面原文/现象 | 能否定点修 |
|---|---|---|---|
| **A. 共享索引页上的“列表项”**（有名字、无独立页、无本单元方向文字） | 5（#2/4/8/19/29） | CWI sub-theme / CLMI theme / ARC theme / CQMT strand / MSRC area / Pharmacy area 都只在父页当 bullet 列出 | **能**：判据加一条——“方向文字须能在页面上落到该单元名下”；不能的降级父单元 |
| **B. 服务/教学/学院级单位** | 5（#5/11/18/20/26） | “support and innovation resource” / “where your future in dentistry begins” / institute | **能**：按 §4「服务设施 · 学生培养计划 · 大一级 institute」剔除 |
| **C. 死链 / URL 指错页** | 2（#14 404、#25 指到索引页） | 404 / 索引页无该条 | **能**：改 URL 或剔除 |

### 4.3 定点修（只修这三类，不逐页重核）

1. **类型 A**：对共享索引页的单元，要求“去掉所有同级单元名后，方向文字仍 ≥60 字”。ARC/CLMI/CQMT/MSRC 的内层 sub-theme 全部不达标 ⇒ **改为只计父中心**（ARC / CQMT / CLMI / MSRC），EEECS 的 CWI/CSIT 同样只计中心。NBE Research Cells、IGFS themes、WWIEM themes、Pharmacy areas 保留（父页对每个单元都有独立方向段）。
2. **类型 B**：剔除 `MethodsLab`、`Centre for Dentistry`、`ARK`、`Northern Ireland Technology Centre`、`Centre for Climate, the Environment and Chronology`、`Brian Friel Theatre` 等服务/教学/设施，以及 `Institute of Irish Studies / Institute of Cognition and Culture / Gibson Institute / William J Clinton Leadership Institute / Institute for Health Sciences / Momentum One Zero` 等大一级 institute（其内层若有具名单元则收，如 IGPSJ 的 8 个 research priority、IGFS 的 4 个 theme）。
3. **类型 C**：修 Pharmacy 15 个 area 的 URL（指向各自 theme 页），删 `InteRRaCt Lab`（404，判「未验」）。

定点修后候选池 204 → **最终 133**（另 71 条落选，逐条原因见 `qub_v2__rows_all.json` 的 `exclude_reason`）。

### 4.4 第二轮抽样 30 条（**对最终 133 条**重新随机抽，逐条开页面核）

| # | 类型 | 名称 | 打开后 page.url()（截断） | h1 · 名称命中 | 判定 |
|---|---|---|---|---|---|
| 1 | cell | Futures Presents Pasts | `/schools/NBE/Research/ResearchCells/` | Research Cells · 3/3 | ✔ |
| 2 | group | Novel Antimicrobials and Anti-infective Biomaterials | `.../ResearchThemes/InfectionandAntimicrobialResistance/` | Infection and Antimicrobial Resistance · 5/5 | ✔ |
| 3 | theme | Rights and Social Justice | `.../ResearchPriorities/RightsandSocialJusticeResearchPriorities/` | Rights and Social Justice Research · 3/3 | ✔ |
| 4 | centre | Precision Medicine Centre of Excellence | `.../cancer-research/CentresofExcellence/PrecisionMedicineCentreofExcellence/` | Precision Medicine Centre of Excellence · 4/4 | ✔ |
| 5 | centre | Centre of Canadian Studies | `.../CentreofCanadianStudiesCCS/` | Centre of Canadian Studies (CCS) · 3/3 | ✔ |
| 6 | group | Health Services and Global Health Research Group | `.../CentreforPublicHealth/Research/HealthServicesGlobalHealth/` | Health Services & Global Health · 6/6 | ✔ |
| 7 | centre | Centre for Data Digitisation and Analysis | `/research-centres/CentreforDataDigitisationandAnalysis/` | Centre for Data Digitisation and Analysis · 4/4 | ✔ |
| 8 | theme | Immunology and Infection Biology | `/research-centres/wwiem/OurResearch/ResearchThemes/` | Research Themes · 3/3 | ✔ |
| 9 | group | Science and Culture Research Group | `/sites/ScienceandCultureResearchGroup/` | Welcome to the Science and Culture RG · 4/4 | ✔ |
| 10 | theme | Nutrition and Preventive Medicine | `/Research/GRI/TheInstituteforGlobalFoodSecurity/research/` | Research · 3/3 | ✔ |
| 11 | theme | Advancing Engineering Education | `.../Research/engineering-education/` | Advancing Engineering Education · 3/3 | ✔ |
| 12 | group | Quality of Care in Vulnerable Populations | `.../ResearchThemes/HealthcareDeliveryandMedicinesOptimisation/` | Healthcare Delivery and Medicines · 4/4 | ✔ |
| 13 | group | Epidemiology and Public Health Research Group | `.../CentreforPublicHealth/Research/EpidemiologyPublicHealth/` | Epidemiology & Public Health · 5/5 | ✔ |
| 14 | group | Brain & Behaviour sub-group | `.../ResearchThemes/DevelopmentCognitionandNeuroscience/` | Development, Cognition and Neuroscience · 3/3 | ✔ |
| 15 | group | Marine Research Group | `.../NBE/Research/facilities-infrastructure/MarineResearchGroup/` | Marine Research Group · 3/3 | ✔ |
| 16 | centre | Centre for Light-Matter Interactions | `/research-centres/light-matter-interactions/` | Centre for Light-Matter Interactions · 4/4 | ✔ |
| 17 | lab | Social Interactions Lab | `.../ResearchThemes/SocialPsychologyandtheStudyofSocialIssues/` | Social Psychology and the Study of Social Issues · 2/2 | ✔ |
| 18 | centre | Centre for Evidence-Based Youth Services | `/research-centres/cebys/` | Centre for Evidence-Based Youth Services · 5/5 | ✔ |
| 19 | group | Respiratory Infectious Diseases | `.../ResearchThemes/InfectionandAntimicrobialResistance/` | Infection and Antimicrobial Resistance · 3/3 | ✔ |
| 20 | centre | Centre for Public Policy and Administration | `/research-centres/cppa/` | Centre for Public Policy and Administration · 4/4 | ✔ |
| 21 | centre | QUOTE | `https://quote.qub.ac.uk/` | (无 h1) · 1/1 | ✔ |
| 22 | group | Breast and Ovarian Cancer | `.../OurResearch/BreastCancerOvarianCancer/` | Breast Cancer & Ovarian Cancer · 3/3 | ✔ |
| 23 | centre | Centre for Creative Ethnography | `/research-centres/CentreforCreativeEthnography/` | About the Centre · 3/3 | ✔ |
| 24 | centre | Centre for Intelligent Sustainable Computing | `/research-centres/cisc/` | Centre for Intelligent Sustainable Computing · 4/4 | ✔ |
| 25 | centre | Centre for Sustainability, Equality and Climate Action | `/research-centres/SECA/` | Centre for Sustainability, Equality and Climate Action · 5/5 | ✔ |
| 26 | theme | Gender, Peace and Security | `.../ResearchPriorities/GenderPeaceandSecurity/` | Gender, Peace and Security · 3/3 | ✔ |
| 27 | centre | Sonic Arts Research Centre | `/research-centres/sarc/` | SARC: Centre for Interdisciplinary Research · 2/4 | ✔ |
| 28 | centre | Centre for Biomedical Sciences Education | `/schools/mdbs/biomedical-sciences-centre/` | Biomedical Sciences Education · 4/4 | ✔ |
| 29 | centre | Animal Behaviour Centre | `/research-centres/animal-behaviour-centre/` | Animal Behaviour Centre · 3/3 | ✔ |
| 30 | centre | Centre for Wireless Innovation | `/research-centres/cwi/` | Wireless Innovation · 3/3 | ✔ |

**第二轮准确率：30 / 30 = 100%**（全部满足：`page.url()` 与目标一致、名称在页面上命中、页面有方向文字）。
`#8/#10/#14/#17/#26` 是共享页上的单元——它们各自在父页上有**独立方向段**，且名称命中，判为真组；`#21` QUOTE 无 H1，用正文方向判定。

**结论（按用户三步走）**：第一轮准确率 58.6% < 90%，但假阳性**类型集中**（A 共享页列表项 / B 服务教学大一级 / C 死链），因此**只定点修这三类，不逐页重核**；修后全量直接用。

---

## 5. 反例 / 判不准 / 未验清单

### 5-1 反例（判它不是组，附页面原文）

| # | 页面 | 判定依据（页面原文） |
|---|---|---|
| 1 | `/research-centres/arc/`（Advanced Research & Engineering Centre） | “was a £9.8 million Invest NI-funded research **partnership** … (2021–2026)” —— 资助/合作项目 |
| 2 | `/research-centres/EPSRCCentreforDoctoralTraininginPhotonicIntegration…/` | “Centre for **Doctoral Training** (CDT) … bringing together postgraduate researchers” —— 学生培养计划 |
| 3 | `/research-centres/nicr/`、`/research-centres/NorthernIrelandCerebralPalsyRegister/` | “a confidential **record** of children with cerebral palsy…” —— 登记册/数据服务 |
| 4 | `/research-centres/PRIMEStudy/`、`/research-centres/YDS/` | “A Longitudinal **Investigation** of Periodontal Disease…” / “Longitudinal **Study** of Adolescent Drug Use” —— 单项研究/队列 |
| 5 | `/research-centres/QueensUniversityMarineLaboratory/`、`/research-centres/TheWellcomeTrust-Wolfson…/`、`/research-centres/QAMEC/` | “a multidisciplinary research **laboratory used by** … staff and students” / “Clinical Research **Facility**” / “Specialist **Technologies / Specialist Facilities**” —— 共享设施 |
| 6 | `/research-centres/ModiolusRestorationResearchGroup/` | “was **commissioned** in 2008 … to provide technical and scientific **support** for the Modiolus Restoration Plan” —— 委托技术支持组 |
| 7 | `/research-centres/SoundingConflict/`、`/research-centres/KnowledgeExchangeSpatialAnalysis…/`、`/research-centres/Remediate/` | 页面自称 “This **project** …” —— 单项项目 |
| 8 | `/schools/ael/Research/ResearchinEnglish/…` 的 `Crossing Frontiers` | “an international **Research Network**” 且非本校挂靠站点 —— 网络 |
| 9 | Law `/research/research-projects/*`（~30 条） | 页面名即 “Research **Project**” —— 单项项目 |
| 10 | NBE `/Research/research-projects/*`（~70 条） | 同上 |
| 11 | `/schools/SchoolofMechanicalandAerospaceEngineering/Research/engineering-education/` 以外的 “Awards/*”、“ConferencesandSymposia/*”、“Public Lectures” | 奖项/会议/讲座系列 |
| 12 | `/schools/psy/Research/ResearchCultureandEnvironment/`、`/schools/ael/Research/OurPostgraduateResearchCommunity/` | 研究环境/PGR 社区，非研究组 |
| 13 | `/schools/InstituteofCognitionCulture/Research/` 的 `Explaining Atheism` 等 | “Current Research **Projects**” —— 项目 |
| 14 | `/Research/GRI/` 的 Momentum One Zero、`/Research/GRI/TheInstituteforHealthSciences/` | 校级研究所（大一级），内层组不足 |
| 15 | `/sites/institute-for-global-peace-security-justice/ResearchandImpact/Blogs/*`（~30 条博客） | 博客文章 |

> 反例共同点仍是那句**「非团队自述」**：`project`、`partnership`、`training`、`registry`、`facility`、`commissioned … support`、`awards`、`blogs`。

### 5-2 判不准（拿不准的放这儿，不自己发明规则归类）

| # | 条目 | 我看到的 / 为什么拿不准 |
|---|---|---|
| 1 | 校级 `Centre for Public Health` / `WWIEM` / `PGJCCR` / `CSIT` / `CWI` / `CLMI` / `ARC` / `CQMT` / `MSRC` | 中心/研究所 vs「导师带的组」；本波按「有内层具名组的，只收内层、不计父」处理，但 CLMI/ARC/CQMT/MSRC 的内层 sub-theme 无独立页无方向文字，只好**计父中心**。严格按“theme 不收”会把这 4 个也去掉、M&P 归 0。 |
| 2 | NBE 的 8 个 Research Cell | 是“staff elect to join”的**社群**（communities of practice），非固定 PI 团队；但原文明确 “Each Research Cell is headed by a Professor or Senior Academic”，故收。 |
| 3 | Pharmacy 的 15 个 area of activity | 每 area 列 2–10 位具名 PI，无单一 lead；是「组」还是「主题分栏」两可。 |
| 4 | MAE 的 6 个 research theme | 有独立页 + mission + funding，但无单一 lead、跨多名 staff；本波保守起见**收**（与曼彻斯特“areas of expertise 计入”的口径一致）。 |
| 5 | Nursing 的 3 个 research theme | 各有 Theme Lead 与团队 ⇒ 收；但名义上仍是 “theme”。 |
| 6 | Law 的 6 个 thematic area | 是 School 内**按主题的 staff 分组**（“Staff research interests are aligned across six thematic areas”），无独立组织实体；本波收（否则 Law 只剩 4 条中心）。 |
| 7 | `QUOTE`、`ARK`、`CEBYS`、`Campbell UK & Ireland` | research resource / policy hub / all-Ireland centre —— 中心偏服务，但都有研究方向与人员自述。ARK 与 14chrono 已按服务剔除，QUOTE/CEBYS/Campbell 保留。 |
| 8 | `Centre for Biomedical Sciences Education` / `Centre for Medical Education` | 名义是 education centre，但属 MDBS 研究结构一层；本波收（边界）。 |
| 9 | `Centre of Canadian Studies`、`Centre for International Borders Research` | 是真中心但规模偏小（1–2 PI）；收。 |
| 10 | `Modiolus Restoration Research Group` | 有 “Research Group” 名与方向，但原文是 commissioned support group；本波**收**（保留在 University Research Centres）。 |

### 5-3 未验（显式列出，不静默跳过）

| # | 条目 | 卡点（URL / 错误原文） |
|---|---|---|
| 1 | InteRRaCt Lab（Psy） | `https://www.qub.ac.uk/schools/psy/Research/ResearchFacilities/` → 实测 **H1 = “PAGE NOT FOUND”**（4 次重试同）。产品中已删；其正确页未找到 |
| 2 | Reimagining Supply Chains Network Plus (RiSC+) | `https://riscplus.com/` → 落在 `/.well-known/sgcaptcha/`（SiteGround captcha），**取不到** |
| 3 | Centre for Theoretical Atomic, Molecular and Optical Physics (CTAMOP) | `http://www.am.qub.ac.uk/ctamop/` → 重定向回 `/schools/SchoolofMathematicsandPhysics/Research/ResearchCentres/`；**已无独立页**，未计入 |
| 4 | Centre for Plasma Physics | 官方 department list 列名，但站内无独立页（M&P ResearchCentres 页只列 4 个）；**未计入** |
| 5 | Atomistic Simulation Research Centre | department list 列名；站内只有 CQMT 的 `research-strands/atomistic-simulations/` 子页；**未计入**（无独立中心页） |
| 6 | QBS `Centre for Economics, Policy and History (CEPH)` | 外站 `ceph.ie` 返回 202 空壳；方向取自 QBS 的 research-centres 索引页 |
| 7 | `Momentum One Zero`（GRI） | `/Research/GRI/` 只给概述，未找到内层具名组页 |
| 8 | School of Biological Sciences 更细的 PI 组 | `/Research/ResearchThemes/` 只到 3 个 theme（各有 Theme Lead），站内无下钻的具名组页；未验“是否还有未公布的 PI 组” |
| 9 | Pure 门户 `pure.qub.ac.uk/en/organisations/` | HTTP 403（Cloudflare JS 挑战）；**未用 Pure 做枚举**（QUB 的组不在 Pure 上组织） |
| 10 | 外站 `crossingfrontiers.wordpress.com` | 未见有效内容页，未计入 |

### 5-4 字段口径说明

- **课题组名称**：取其**自己页面**的 H1；共享父页上的单元用父页里的官方名（`page_kind = parent index (shared URL)`，共 41 条）。
- **所在大学及研究中心**：只有上级是**真正的 research centre/institute** 才写 `Queen's University Belfast | <unit>`（CPH、PGJCCR、WWIEM、CWI、CSIT）；上级是 School/Department 的一律**只写大学**。
- **研究方向**：取该组页的方向文字（去掉导航与 “Link -” 后 ≥60 字）；**没有 ⇒ 留空**（本波 133 条全部有方向）。
- **联系邮箱**：只接受**本组页**上的邮箱；本波 63/133 条在组页上有邮箱，其余留空（未去别处凑字段）。
- **最近五年论文数**：133/133 = `null`（用户裁定）。

---

## 6. 一句话结论

**贝尔法斯特女王大学最终组数 133（15 个官方 School 下 117 条 + 研究所/校级中心具名内层单元 16 条），组/系 ≈8.9；完备性已从上一轮“只看校级中心总表”的 29 组补齐到逐 School research 子树 + 中心内层具名组 + `/sites/` PI 组站级别；剩余缺口只有四处且都写明了卡点——M&P 的 CTAMOP/Centre for Plasma Physics 无独立页、Biological Sciences 只有 3 个 theme 无下钻组、InteRRaCt Lab 页 404、RiSC+ 被 captcha 挡。**