Saved in:
Bibliographic Details
Main Authors: Pavlova, Vera, Makhlouf, Mohammed
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.16797
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910004035977216
author Pavlova, Vera
Makhlouf, Mohammed
author_facet Pavlova, Vera
Makhlouf, Mohammed
contents We introduce MOSAIC (Masked Objective with Selective Adaptation for In-domain Contrastive learning), a multi-stage framework for domain adaptation of text embedding models that incorporates joint domain-specific masked supervision. Our approach addresses the challenges of adapting large-scale general-domain text embedding models to specialized domains. By jointly optimizing masked language modeling (MLM) and contrastive objectives within a unified training pipeline, our method enables effective learning of domain-relevant representations while preserving the robust semantic discrimination properties of the original model. We empirically validate our approach on both high-resource and low-resource domains, achieving improvements up to 13.4% in NDCG@10 (Normalized Discounted Cumulative Gain) over strong general-domain baselines. Comprehensive ablation studies further demonstrate the effectiveness of each component, highlighting the importance of balanced joint supervision and staged adaptation.
format Preprint
id arxiv_https___arxiv_org_abs_2510_16797
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MOSAIC: Masked Objective with Selective Adaptation for In-domain Contrastive Learning
Pavlova, Vera
Makhlouf, Mohammed
Computation and Language
Artificial Intelligence
We introduce MOSAIC (Masked Objective with Selective Adaptation for In-domain Contrastive learning), a multi-stage framework for domain adaptation of text embedding models that incorporates joint domain-specific masked supervision. Our approach addresses the challenges of adapting large-scale general-domain text embedding models to specialized domains. By jointly optimizing masked language modeling (MLM) and contrastive objectives within a unified training pipeline, our method enables effective learning of domain-relevant representations while preserving the robust semantic discrimination properties of the original model. We empirically validate our approach on both high-resource and low-resource domains, achieving improvements up to 13.4% in NDCG@10 (Normalized Discounted Cumulative Gain) over strong general-domain baselines. Comprehensive ablation studies further demonstrate the effectiveness of each component, highlighting the importance of balanced joint supervision and staged adaptation.
title MOSAIC: Masked Objective with Selective Adaptation for In-domain Contrastive Learning
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2510.16797