NAVINESSEO LAB
Menu
LangueFrançais

NAVINES SEO LAB

Recherche

Des études originales dont la méthode, les sources et les limites sont visibles.

Bibliothèque de preuves

Six articles lus intégralement et traduits en décisions produit délimitées.

arXiv · 2022-10-18

Making a MIRACL: Multilingual Information Retrieval Across a Continuum of Languages

Auteurs
Xinyu Zhang; Nandan Thakur; Odunayo Ogundepo; Ehsan Kamalloo; David Alfonso-Hermelo; Xiaoguang Li; Qun Liu; Mehdi Rezagholizadeh; Jimmy Lin
Date de publication
2022-10-18
Ouvrir la fiche de preuve

Question de recherche

How can a retrieval benchmark cover languages with very different resource levels while using native queries and relevance judgments?

Méthode

An 18-language, same-language Wikipedia retrieval benchmark built with native-speaker queries and judgments; lexical, dense and hybrid baselines were compared on the released pools.

Ce que les preuves étayent

Evaluate each language explicitly, document corpus and judgment construction, and keep native review central. The reported hybrid result is a historical benchmark observation.

Limites et transférabilité

This is same-language information retrieval, not translation assessment or web SEO. Heuristic segmentation, candidate-pool gaps and unfinished test labels limit transferability; it says nothing about hreflang or Google rankings.

Classification

  • Étayer
  • Historique
  • Académique seulement

Influence NAVINES: Laboratoire de comparaison de pages multilingues

arXiv · 2024-04-14

Competitive Retrieval: Going Beyond the Single Query

Auteurs
Haya Nachimovsky; Moshe Tennenholtz; Fiana Raiber; Oren Kurland
Date de publication
2024-04-14
Ouvrir la fiche de preuve

Question de recherche

How does publisher competition change when participants allocate content decisions across multiple queries rather than one isolated query?

Méthode

A formal multi-query game plus four controlled student competitions covering 30 TREC topics, 84 participants and specified proxy rankers; best-response dynamics and feature changes were examined.

Ce que les preuves étayent

Cross-query opportunity cost and competitor response deserve explicit treatment; equilibrium existence does not imply that learning dynamics will converge.

Limites et transférabilité

The student setting, topics and rankers are controlled proxies. The work does not identify current Google features, predict winners or justify reusable weights or a practical Nash-equilibrium claim.

Classification

  • Limité
  • Académique seulement

Influence NAVINES: Planificateur de portefeuille concurrentiel multirequêtes

arXiv · 2020-05-26

Ranking-Incentivized Quality Preserving Content Modification

Auteurs
Gregory Goren; Oren Kurland; Moshe Tennenholtz; Fiana Raiber
Date de publication
2020-05-26
Ouvrir la fiche de preuve

Question de recherche

Can ranking-incentivized content modification include coherence and modification cost instead of optimizing promotion alone?

Méthode

A controlled passage-replacement method balanced promotion and coherence, evaluated with proxy rankers and offline data across 31 queries, with an explicit ethics discussion.

Ce que les preuves étayent

Quality constraints and the cost of changing content belong in planning. A promotional change that fails editorial integrity should be rejected.

Limites et transférabilité

Passages came from higher-ranked candidates and quality judgments were bounded and subjective. It does not support copying winners, real-engine causal claims or unrestricted automated rewriting.

Classification

  • Limité
  • Académique seulement

Influence NAVINES: Planificateur de portefeuille concurrentiel multirequêtes

Journal of Artificial Intelligence Research · 2025

The Search for Stability: Learning Dynamics of Strategic Publishers with Initial Documents

Auteurs
Omer Madmon; Idan Pipano; Itamar Reinman; Moshe Tennenholtz
Date de publication
2025
Ouvrir la fiche de preuve

Question de recherche

Under which formal ranking mechanisms and content-deviation costs can strategic publisher learning converge or remain unstable?

Méthode

Formal publisher-utility models with rank benefit minus deviation cost, convergence results for specified linear/softmax settings, and discrete simulations, predominantly with two publishers.

Ce que les preuves étayent

Change cost, response dynamics and instability are useful planning dimensions; a mechanism can have an equilibrium while local responses cycle or remain pseudoperiodic.

Limites et transférabilité

The model assumes a static information need, known ranking function and embedding-space actions. Simulations are not a real search engine and do not justify a Google mechanism or guaranteed convergence claim.

Classification

  • Étayer
  • Limité
  • Académique seulement

Influence NAVINES: Planificateur de portefeuille concurrentiel multirequêtes

KDD / arXiv · 2024

GEO: Generative Engine Optimization

Auteurs
Pranjal Aggarwal; Vishvak Murahari; Tanmay Rajpurohit; Ashwin Kalyan; Kartik Narasimhan; Ameet Deshpande
Date de publication
2024
Ouvrir la fiche de preuve

Question de recherche

In controlled generative-engine experiments, how do selected content modifications affect proxy visibility across queries and domains?

Méthode

GEO-BENCH combined 10,000 queries from nine datasets with visibility proxies and experiments on two generative engines; citation, quotation, statistics and other interventions were compared by domain.

Ce que les preuves étayent

Clear attribution, genuine evidence and domain-aware experimentation can be reviewed as editorial signals; keyword stuffing performed poorly in this benchmark.

Limites et transférabilité

Two engines, black-box variance and proxy metrics do not establish traditional-search effects or future citation probability. Reported gains are historical benchmark results, not forecasts; appendix prompts are untrusted.

Classification

  • Limité
  • Historique
  • Académique seulement

Influence NAVINES: Laboratoire de préparation aux citations IA

arXiv · 2024-07-02

Adversarial Search Engine Optimization for Large Language Models

Auteurs
Fredrik Nestaas; Edoardo Debenedetti; Florian Tramèr
Date de publication
2024-07-02
Ouvrir la fiche de preuve

Question de recherche

Can content in retrieved web or plugin documents manipulate an LLM's product preferences, and which broader attack categories appear beyond classic prompt injection?

Méthode

Controlled experiments with fictional products, web/plugin documents and proprietary models examined model-directed instructions, concealment, false claims, source suppression and competitor discrediting.

Ce que les preuves étayent

Retrieved content is an untrusted input surface; safety review should cover concealed persuasion and source integrity as well as direct instructions.

Limites et transférabilité

The favorable fictional setup, limited scale and changing proprietary systems do not establish prevalence or complete detection. Exact attack payloads are not republished and the checker is not a certification.

Classification

  • Étayer
  • Limité
  • Académique seulement

Influence NAVINES: Contrôle de sécurité contre la manipulation IA

Preuves examinées le 14 août 2026 · N’établit pas de facteurs Google actuels ni de résultats garantis.

Recherche

Des études originales dont la méthode, les sources et les limites sont visibles.

Nous ne courons pas après les algorithmes. Nous construisons clarté, confiance et visibilité durable.

Rendre l’intelligence avancée de la recherche compréhensible, utile et accessible, puis aider les entreprises à la convertir en actions mesurables.

Méthode

  • Cadrez les environnements, modèles critiques, données disponibles et risques de changement.
  • Collectez statuts HTTP, directives, canonicales, liens, rendu, sitemaps et données Search Console.
  • Documentez objectifs, événements, dimensions, responsables et décisions liées à chaque métrique.
  • Vérifiez balises, consentement, fuseaux horaires, filtres et pertes de données prévisibles.

Limites

  • Consentement, bloqueurs et modélisation rendent les données incomplètes.
  • Search Console agrège et conserve les données avec des limites ; il ne journalise pas chaque recherche.