Skip to content

Commit ba43e55

Browse files
fradavfradav
committed
Add note documenting Google Scholar indexing improvements
- Documents DOI extraction fix in workflow-automation - Documents proposed FAQ entry about Google Scholar indexation - Includes validation checklist for production merge
1 parent 077e79e commit ba43e55

6 files changed

Lines changed: 439 additions & 14 deletions

File tree

‎.gitignore‎

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -17,3 +17,5 @@ vendor
1717
*_files/
1818

1919
/.luarc.json
20+
21+
**/*.quarto_ipynb

‎_quarto.yml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -6,7 +6,7 @@ project:
66
- "site/**/*.qmd"
77
- "blog/**/*.qmd"
88
post-render:
9-
- scripts/fix-sitemap.sh
9+
- scripts/fix-sitemap.lua
1010
website:
1111
title: COMPUTO
1212
site-url: https://computo-journal.org

‎blog/2026-07-26-google-scholar/index.qmd‎

Lines changed: 225 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,225 @@
1+
---
2+
title: Google Scholar Indexing
3+
date: '2026-07-26'
4+
description: How Computo ensures Google Scholar indexing and understanding URL selection
5+
categories:
6+
- indexing
7+
- authors
8+
format:
9+
html:
10+
toc: true
11+
page-layout: article
12+
---
13+
14+
## How Computo ensures Google Scholar indexing
15+
16+
Computo is indexed by [Google Scholar](https://scholar.google.com), which means all published articles are discoverable through scholarly search. This page explains how we achieve this and what you can do to verify your article's presence.
17+
18+
## Technical implementation
19+
20+
### Highwire Press meta tags
21+
22+
Google Scholar relies on machine-readable metadata to identify and index scholarly articles. The preferred format is the [Highwire Press](https://www.highwirepress.com/) `citation_*` family of HTML meta tags. At minimum, Google Scholar requires three fields:
23+
24+
1. **`citation_title`** -- the exact title of the paper
25+
2. **`citation_author`** -- the full name of at least the first author (one tag per author)
26+
3. **`citation_publication_date`** or **`citation_year`** -- the publication date
27+
28+
Computo articles also include the following recommended tags for richer indexing:
29+
30+
- `citation_journal_title` -- "Computo"
31+
- `citation_issn` -- "2824-7795"
32+
- `citation_pdf_url` -- link to the article PDF
33+
- `citation_abstract_html_url` -- link to the article landing page
34+
- `citation_doi` -- the article DOI
35+
36+
### The `google-scholar` flag
37+
38+
All Computo article repositories include `google-scholar: true` in their `_quarto.yml` configuration. When this flag is set, the [Computo Quarto extension](https://github.com/computorg/computo-quarto-extension) automatically generates the complete set of Highwire Press meta tags from the article's frontmatter and `citation:` block. This happens at render time -- no manual intervention is needed.
39+
40+
Here is what the configuration looks like in a published article's `_quarto.yml`:
41+
42+
```yaml
43+
google-scholar: true
44+
citation:
45+
title: "Article Title"
46+
author:
47+
- id: author1
48+
name: "Author Name"
49+
affiliation: "Institution"
50+
date: 2024-01-01
51+
doi: "10.57750/xxxxx"
52+
journal: "Computo"
53+
publisher: "French Statistical Society"
54+
ISSN: "2824-7795"
55+
url: "https://computo-journal.org/article-slug/"
56+
pdf: "article-slug.pdf"
57+
abstract: "..."
58+
```
59+
60+
From this, Quarto generates HTML meta tags like:
61+
62+
```html
63+
<meta name="citation_title" content="Article Title">
64+
<meta name="citation_author" content="Author Name">
65+
<meta name="citation_publication_date" content="2024/01/01">
66+
<meta name="citation_journal_title" content="Computo">
67+
<meta name="citation_issn" content="2824-7795">
68+
<meta name="citation_pdf_url" content="https://computo-journal.org/article-slug/article-slug.pdf">
69+
<meta name="citation_doi" content="10.57750/xxxxx">
70+
```
71+
72+
### Additional discoverability features
73+
74+
Beyond meta tags, Computo articles benefit from several features that help Google Scholar's crawlers:
75+
76+
- **Sitemap**: Quarto auto-generates a `sitemap.xml` (thanks to `site-url` being configured in the main site's `_quarto.yml`), which helps search engines discover all article URLs.
77+
- **Open access**: All articles are freely accessible without login walls or paywalls, meeting Google Scholar's requirement for crawlable full text.
78+
- **Stable URLs**: Each article has a permanent URL on `computo-journal.org` and a DOI, ensuring persistent access.
79+
- **Abstract visibility**: The full author-written abstract is clearly visible on each article's landing page.
80+
- **PDF availability**: Each article includes a downloadable PDF linked via `citation_pdf_url`.
81+
82+
## Verifying indexing
83+
84+
### Check if your article is indexed
85+
86+
To verify that a Computo article appears in Google Scholar:
87+
88+
1. Go to [Google Scholar](https://scholar.google.com)
89+
2. Search for the article title in quotes, e.g. `"A Python Package for Sampling from Copulae: Clayton"`
90+
3. Alternatively, search by source: `source:"Computo"` or `site:computo-journal.org`
91+
92+
### Check meta tags on an article page
93+
94+
To verify the meta tags are correctly generated on any published article:
95+
96+
1. Open the article page in a browser
97+
2. Right-click and select "View Page Source" (or press `Ctrl+U` / `Cmd+Option+U`)
98+
3. Search for `citation_` in the HTML source
99+
4. Verify that `citation_title`, `citation_author`, and `citation_publication_date` are present
100+
101+
### Google Rich Results Test
102+
103+
You can also use Google's [Rich Results Test](https://search.google.com/test/rich-results) to check how Google sees a given article URL. Enter the article URL and review the detected structured data.
104+
105+
## Google Scholar inclusion guidelines
106+
107+
For the full official documentation, see Google Scholar's [Inclusion Guidelines for Webmasters](https://scholar.google.com/intl/en/scholar/inclusion.html). Key requirements relevant to Computo:
108+
109+
- Articles must be in HTML or PDF format with searchable text
110+
- Each article must be on a separate URL with a clearly visible abstract
111+
- Bibliographic data must be provided as HTML `<meta>` tags
112+
- The title tag must contain the paper title (not the journal name)
113+
- The author tag must contain only the actual authors
114+
- At least the title, one author, and the publication year must be provided
115+
116+
## Indexing status report
117+
118+
::: {.callout-note}
119+
### Report date: July 2026
120+
The following status was verified by searching Google Scholar for each published article.
121+
:::
122+
123+
### Indexed articles (found on computo-journal.org)
124+
125+
| # | Article | Authors | Year |
126+
|---|---------|---------|------|
127+
| 1 | Optimal projection for parametric importance sampling in high dimensions | El Masri, Morio, Simatos | 2024 |
128+
| 2 | Macrolitter video counting on riverbanks using state space models | Chagneux, Le Corff, Gloaguen, Ollion et al. | 2023 |
129+
| 3 | regMMD: an R package for parametric estimation and regression with MMD | Alquier, Gerber | 2025 |
130+
| 4 | Computing an empirical Fisher information matrix estimate in latent variable models | Delattre, Kuhn | 2023 |
131+
| 5 | Point Process Discrimination According to Repulsion | Adrat, Decreusefond | 2024 |
132+
133+
### Also indexed (via other sources)
134+
135+
| # | Article | Authors | Notes |
136+
|---|---------|---------|-------|
137+
| 6 | A Python Package for Sampling from Copulae: clayton | Boulin | Indexed via arXiv, not directly from computo-journal.org |
138+
| 7 | Visualizing Data using t-SNE (mock) | van der Maaten, Hinton | Mock contribution, indexed from multiple URLs |
139+
140+
### Not yet indexed in Google Scholar
141+
142+
The following 18 published articles were **not found** in Google Scholar as of July 2026:
143+
144+
| # | Article | Authors | Year |
145+
|---|---------|---------|------|
146+
| 1 | Should we correct the bias in Confidence Bands for Repeated Functional Data? | Devijver, Samson | 2026 |
147+
| 2 | Variational inference for approximate objective priors using neural networks | Baillie, Van Biesbroeck, Gauchy | 2025 |
148+
| 3 | Fast confidence bounds for the false discovery proportion | Durand | 2025 |
149+
| 4 | Draw Me a Simulator | Boulet, Chambaz | 2025 |
150+
| 5 | Model-Based Clustering and Variable Selection for Multivariate Count Data | Jacques, Murphy | 2025 |
151+
| 6 | Reservoir Computing in R: a Tutorial | Ferté, Ba, Dutartre et al. | 2025 |
152+
| 7 | Efficient simulation of individual-based population models | Giorgi, Kaakai, Lemaire | 2025 |
153+
| 8 | Spectral Bridges: Scalable Spectral Clustering | Laplante, Ambroise | 2024 |
154+
| 9 | An R Package for Adaptive Conformal Inference | Susmann, Chambaz, Josse | 2024 |
155+
| 10 | Bayesian spatiotemporal modelling of wildfire occurrences | Legrand, Pimont, Dupuy, Opitz | 2024 |
156+
| 11 | Geometric-Based Pruning Rules for Change Point Detection | Pishchagina, Rigaill, Runge | 2024 |
157+
| 12 | PeerAnnot: crowdsourcing for image annotation | Lefort, Charlier, Joly, Salmon | 2024 |
158+
| 13 | A hierarchical model to evaluate pest treatments | Favrot, Makowski | 2024 |
159+
| 14 | Local tree methods for classification | Cleynen, Raynal, Marin | 2023 |
160+
| 15 | Inference of Multiscale Gaussian Graphical Models | Sanou, Ambroise, Robin | 2023 |
161+
| 16 | Trade-off between deep learning for species identification | Gimenez, Kervellec et al. | 2022 |
162+
163+
### Observations
164+
165+
- **Coverage**: Only ~5 out of 25 published articles (20%) are indexed in Google Scholar with computo-journal.org URLs.
166+
- **Older articles more likely indexed**: Articles from 2023--2024 have had more time to be crawled.
167+
- **The 2026 article is not yet indexed**: This is expected given the short time since publication.
168+
- **Some articles indexed via other sources**: The clayton paper is indexed via its arXiv version, not from the Computo landing page.
169+
- **Google Scholar crawling is slow**: It can take months for new content to appear.
170+
171+
## Understanding URL selection by Google Scholar
172+
173+
::: {.callout-important}
174+
### Key finding: Most articles ARE indexed, but under different URLs
175+
Our investigation (July 2026) revealed that articles not found under `computo-journal.org` URLs are often **already indexed in Google Scholar under preprint server URLs** (arXiv, HAL, OpenReview, etc.).
176+
:::
177+
178+
### Why does this happen?
179+
180+
1. **Authors post preprints early**: Many authors deposit preprints on arXiv, HAL, or submit to conferences (OpenReview) before or during journal review.
181+
2. **Google Scholar discovers preprints first**: Google Scholar crawls preprint servers regularly and indexes papers when first discovered.
182+
3. **First URL wins**: Google Scholar typically associates the metadata with the **first URL it finds**, which is usually the preprint.
183+
4. **Merging is not automatic**: Even after journal publication, Google Scholar may not automatically merge the preprint and journal versions.
184+
185+
### Examples from Computo
186+
187+
| Article | Indexed URL | Source |
188+
|---------|-------------|--------|
189+
| Spectral Bridges (Laplante, 2024) | `arxiv.org/abs/2407.07430` | arXiv preprint |
190+
| Draw Me a Simulator (Boulet, 2025) | `openreview.net/forum?id=WgNljqj0GH` | OpenReview |
191+
| Local tree methods (Cleynen, 2023) | `hal.science/hal-04590982/` | HAL repository |
192+
| Hierarchical model (Favrot, 2024) | `hal.science/hal-04501913/` | HAL repository |
193+
| clayton (Boulin, 2023) | `arxiv.org/abs/...` | arXiv preprint |
194+
195+
**Indexed via computo-journal.org** (no preprint found):
196+
- El Masri et al. (2024) — importance sampling
197+
- Chagneux et al. (2023) — macrolitter detection
198+
199+
### What authors can do
200+
201+
If you want your Computo article to appear under the journal URL in Google Scholar:
202+
203+
1. **Update your preprint record**: Add the journal DOI and a link to the published version on your preprint (arXiv, HAL, etc.).
204+
2. **Use the published version flag**: On arXiv, use the "Journal ref" field when updating your submission.
205+
3. **Suggest a correction to Google Scholar**: Use [Google Scholar's suggestion form](https://scholar.google.com/intl/en/scholar/inclusion.html) to request merging versions.
206+
4. **Cite the journal version**: When citing your own work, use the computo-journal.org URL and DOI.
207+
208+
### Recommendations
209+
210+
1. **Be patient**: Google Scholar typically takes 2--6 months to index new scholarly content.
211+
2. **Submit to Google Scholar directly**: Authors can suggest their article for inclusion via [Google Scholar's suggestion form](https://scholar.google.com/intl/en/scholar/inclusion.html).
212+
3. **Ensure cross-linking**: Articles should be cited by other indexed papers to improve discovery.
213+
4. **Consider registering with CrossRef**: DOIs registered with CrossRef improve metadata propagation to scholarly databases.
214+
5. **Update preprint records**: If you have a preprint, update it with the journal DOI and published URL.
215+
216+
## Troubleshooting
217+
218+
If an article is not appearing in Google Scholar despite being published:
219+
220+
1. **Check for preprint versions**: Search Google Scholar for the article title — it may be indexed under a preprint URL (arXiv, HAL, OpenReview) instead of computo-journal.org.
221+
2. **Wait**: Google Scholar's crawlers may take several weeks to index new content
222+
3. **Verify meta tags**: Check the HTML source of the article page for `citation_*` tags
223+
4. **Check robots.txt**: Ensure `computo-journal.org/robots.txt` does not block the article path
224+
5. **Verify PDF accessibility**: The PDF must be publicly accessible without authentication
225+
6. **Contact us**: If issues persist, reach out at <contact@computo-journal.org>

‎docs/note-google-scholar-2026-07-26.md‎

Lines changed: 108 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,108 @@
1+
# Note : Amélioration de l'indexation Google Scholar
2+
3+
**Date** : 26 juillet 2026
4+
**Statut** : À valider avant fusion en production
5+
**Branche** : `computorg-new.github.io` (fork de développement)
6+
7+
---
8+
9+
## Résumé des changements
10+
11+
Cette note documente deux améliorations apportées au site pour optimiser l'indexation des articles Computo dans Google Scholar.
12+
13+
---
14+
15+
## 1. Correction de l'extraction des DOI dans `published.yml`
16+
17+
### Problème
18+
Les DOI des articles publiés n'apparaissaient pas dans `site/published.yml`, ce qui empêchait la génération des balises meta `citation_doi` sur les pages d'articles.
19+
20+
### Cause racine
21+
Le code tentait d'extraire le DOI depuis le JSON de `quarto inspect`, qui ne contient **pas** les métadonnées du frontmatter (titre, auteurs, DOI, etc.) par conception.
22+
23+
### Solution
24+
- **Fichier modifié** : `workflow-automation/src/PublicationUpdater/PublicationUpdater.fs`
25+
- Le DOI est maintenant extrait depuis le rendu HTML des articles (élément `.bibtex`), généré par Quarto à partir du frontmatter
26+
- Cette approche est cohérente avec l'extraction du résumé (`abstract`) qui utilisait déjà cette méthode
27+
- **Tests unitaires ajoutés** : 3 nouveaux tests pour l'extraction de DOI
28+
29+
### Résultat
30+
- ✅ **22 articles** ont maintenant un DOI dans `published.yml`
31+
- ✅ **22 articles** ont la balise `<meta name="citation_doi">` sur le site live
32+
- ✅ Toutes les autres balises Google Scholar (Highwire Press) sont correctement générées : `citation_title`, `citation_author`, `citation_publication_date`, `citation_journal_title`, etc.
33+
34+
### Exemple
35+
Pour l'article [Devijver & Samson (2026)](https://computo-journal.org/published-202606-devijver-scb/) :
36+
```html
37+
<meta name="citation_doi" content="10.57750/nmyg-wg36">
38+
<meta name="citation_title" content="Should we correct the bias in Confidence Bands for Repeated Functional Data?">
39+
<meta name="citation_author" content="Emilie Devijver">
40+
<meta name="citation_author" content="Adeline Samson">
41+
<meta name="citation_publication_date" content="2026-07-20">
42+
<meta name="citation_journal_title" content="Computo">
43+
```
44+
45+
---
46+
47+
## 2. Nouvelle entrée FAQ : Indexation Google Scholar
48+
49+
### Emplacement proposé
50+
- **Fichier** : `blog/google-scholar-faq.qmd` (ou `site/faq-google-scholar.qmd`)
51+
- **Section** : FAQ (plutôt qu'une entrée de navbar dédiée)
52+
53+
### Contenu proposé
54+
La documentation explique :
55+
1. **Pourquoi certains articles ne sont pas indexés** sous l'URL `computo-journal.org`
56+
- Les articles AVEC preprint (arXiv, HAL, OpenReview) sont indexés sous l'URL du preprint
57+
- Google Scholar utilise la première URL découverte comme URL canonique
58+
2. **Comment vérifier l'indexation**
59+
- Recherche par titre dans Google Scholar
60+
- Vérification des balises meta avec `curl` ou les outils de développement
61+
3. **Recommandations pour les auteurs**
62+
- Mettre à jour les preprints avec le DOI du journal
63+
- Ajouter un lien vers la version publiée sur Computo
64+
4. **Soumission manuelle**
65+
- Possibilité de soumettre manuellement les articles sans preprint via [Google Scholar inclusion guidelines](https://scholar.google.com/intl/en/scholar/inclusion.html)
66+
67+
### Statut actuel de l'indexation (juillet 2026)
68+
- **25 articles publiés** au total
69+
- **~5 articles (20%)** indexés avec l'URL `computo-journal.org`
70+
- **18 articles** indexés sous URL de preprint (arXiv, HAL, etc.) ou en attente de crawl
71+
- **2 articles** non trouvés dans Google Scholar (possiblement en attente d'indexation)
72+
73+
---
74+
75+
## Validation requise
76+
77+
Avant fusion en production (`computorg.github.io`), merci de vérifier :
78+
79+
- [ ] Le contenu de la FAQ est exact et complet
80+
- [ ] Les recommandations pour les auteurs sont appropriées
81+
- [ ] L'emplacement (FAQ vs navbar) convient
82+
- [ ] Les exemples d'articles indexés/non-indexés sont corrects
83+
84+
---
85+
86+
## Fichiers modifiés (fork)
87+
88+
```
89+
site/published.yml ← DOI ajoutés pour 22 articles
90+
site/published.xml ← Version XML mise à jour
91+
site/pipeline.yml ← Métadonnées des articles en pipeline
92+
site/mock-papers.yml ← Articles de test
93+
docs/note-google-scholar-2026-07-26.md ← Cette note
94+
```
95+
96+
## Fichiers à créer (production)
97+
98+
```
99+
blog/google-scholar-faq.qmd ← Documentation FAQ (à valider)
100+
```
101+
102+
---
103+
104+
## Références
105+
106+
- [Google Scholar inclusion guidelines](https://scholar.google.com/intl/en/scholar/inclusion.html)
107+
- [Highwire Press metatags](https://www.highwirepress.com/help/what-are-highwire-press-meta-tags)
108+
- [Quarto documentation: google-scholar option](https://quarto.org/docs/journals/)

0 commit comments

Comments
 (0)