Integrating Relevance Feedback for Effective Personalisation in
Vector Search

D 2025

Integrating Relevance Feedback for Effective Personalisation in Vector Search

ŠIKYŇA, Matúš a Pavel ZEZULA

Základní údaje

Originální název

Integrating Relevance Feedback for Effective Personalisation in Vector Search

Autoři

ŠIKYŇA, Matúš a Pavel ZEZULA

Vydání

Cham, 18th International Conference on Similarity Search and Applications (SISAP 2025), od s. 154-162, 9 s. 2025

Nakladatel

Springer

Další údaje

Jazyk

angličtina

Typ výsledku

Stať ve sborníku

Obor

10200 1.2 Computer and information sciences

Stát vydavatele

Švýcarsko

Utajení

není předmětem státního či obchodního tajemství

Forma vydání

tištěná verze "print"

Impakt faktor

Impact factor: 0.402 v roce 2005

Označené pro přenos do RIV

Ano

Organizační jednotka

Fakulta informatiky

ISBN

978-3-032-06068-6

ISSN

Klíčová slova anglicky

Similarity search; Personalized similarity; Vector databases; Relevance feedback

Štítky

core_B, DISA, firank_B, similarity search

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 15. 9. 2026 06:51, RNDr. Mgr. Matúš Šikyňa

Anotace

V originále

Vector search systems typically rely on fixed indexing methods and standard similarity measures, such as Euclidean distance, assuming a universal notion of semantic similarity. This assumption overlooks the subjective and context-dependent nature of human similarity perception. Prior approaches proposed personalised querying via Mahalanobis distance parameterised by learned user-profile matrices, but did not address strategies for acquiring real user relevance feedback. We present an end-to-end pipeline designed to integrate semantic feedback directly from user interactions. We explore three feedback strategies, evaluating them across ten established metric learning models. Through an empirical analysis, we assess how these strategies affect key performance indicators, such as retrieval precision. We also outline three practical operational scenarios -- balancing precision and computational costs, prioritising computational efficiency, and optimising for precision -- and provide concrete model and system configuration recommendations. Experiments show that the recommended setups make search efficient while improving retrieval precision, showing the practicality of the approach.

Návaznosti

MUNI/A/1638/2024, interní kód MU

Název: Umělá inteligence a správa komplexních rozsáhlých dat

Investor: Masarykova univerzita, Umělá inteligence a správa komplexních rozsáhlých dat

VK01010147, projekt VaV

Název: Automatizovaná forenzní laboratoř digitálních dat pro odhalování komplexní trestné činnosti

Investor: Ministerstvo vnitra ČR, Automatizovaná forenzní laboratoř digitálních dat pro odhalování komplexní trestné činnosti

Citovat

ŠIKYŇA, Matúš a Pavel ZEZULA. Integrating Relevance Feedback for Effective Personalisation in Vector Search. In 18th International Conference on Similarity Search and Applications (SISAP 2025). Cham: Springer, 2025, s. 154-162. ISBN 978-3-032-06068-6. Dostupné z: https://doi.org/10.1007/978-3-032-06069-3_13.

@inproceedings{2510557,
   author = {Šikyňa, Matúš and Zezula, Pavel},
   address = {Cham},
   booktitle = {18th International Conference on Similarity Search and Applications (SISAP 2025)},
   doi = {https://doi.org/10.1007/978-3-032-06069-3_13},
   keywords = {Similarity search; Personalized similarity; Vector databases; Relevance feedback},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Cham},
   isbn = {978-3-032-06068-6},
   pages = {154-162},
   publisher = {Springer},
   title = {Integrating Relevance Feedback for Effective Personalisation in Vector Search},
   year = {2025}
}

TY  - CONF
ID  - 2510557
AU  - Šikyňa, Matúš - Zezula, Pavel
PY  - 2025
TI  - Integrating Relevance Feedback for Effective Personalisation in Vector Search
PB  - Springer
CY  - Cham
SN  - 9783032060686
KW  - Similarity search
KW  - Personalized similarity
KW  - Vector databases
KW  - Relevance feedback
N2  - Vector search systems typically rely on fixed indexing methods and standard similarity measures, such as Euclidean distance, assuming a universal notion of semantic similarity. This assumption overlooks the subjective and context-dependent nature of human similarity perception. Prior approaches proposed personalised querying via Mahalanobis distance parameterised by learned user-profile matrices, but did not address strategies for acquiring real user relevance feedback. We present an end-to-end pipeline designed to integrate semantic feedback directly from user interactions. We explore three feedback strategies, evaluating them across ten established metric learning models. Through an empirical analysis, we assess how these strategies affect key performance indicators, such as retrieval precision. We also outline three practical operational scenarios -- balancing precision and computational costs, prioritising computational efficiency, and optimising for precision -- and provide concrete model and system configuration recommendations. Experiments show that the recommended setups make search efficient while improving retrieval precision, showing the practicality of the approach.
ER  -

ŠIKYŇA, Matúš a Pavel ZEZULA. Integrating Relevance Feedback for Effective Personalisation in Vector Search. In \textit{18th International Conference on Similarity Search and Applications (SISAP 2025)}. Cham: Springer, 2025, s.~154-162. ISBN~978-3-032-06068-6. Dostupné z: https://doi.org/10.1007/978-3-032-06069-3\_{}13.

Přehled o publikaci