Saved in:
Bibliographic Details
Main Authors: Tiwari, Ashish, Singh, Mukul, Singha, Ananya, Radhakrishna, Arjun
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.10698
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910875148877824
author Tiwari, Ashish
Singh, Mukul
Singha, Ananya
Radhakrishna, Arjun
author_facet Tiwari, Ashish
Singh, Mukul
Singha, Ananya
Radhakrishna, Arjun
contents The goal of diversity sampling is to select a representative subset of data in a way that maximizes information contained in the subset while keeping its cardinality small. We introduce the ordered diverse sampling problem based on a new metric that measures the diversity in an ordered list of samples. We present a novel approach for generating ordered diverse samples for textual data that uses principal components on the embedding vectors. The proposed approach is simple and compared with existing approaches using the new metric. We transform standard text classification benchmarks into benchmarks for ordered diverse sampling. Our empirical evaluation shows that prevailing approaches perform 6% to 61% worse than our method while also being more time inefficient. Ablation studies show how the parts of the new approach contribute to the overall metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2503_10698
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Ordered Semantically Diverse Sampling for Textual Data
Tiwari, Ashish
Singh, Mukul
Singha, Ananya
Radhakrishna, Arjun
Computation and Language
Information Retrieval
The goal of diversity sampling is to select a representative subset of data in a way that maximizes information contained in the subset while keeping its cardinality small. We introduce the ordered diverse sampling problem based on a new metric that measures the diversity in an ordered list of samples. We present a novel approach for generating ordered diverse samples for textual data that uses principal components on the embedding vectors. The proposed approach is simple and compared with existing approaches using the new metric. We transform standard text classification benchmarks into benchmarks for ordered diverse sampling. Our empirical evaluation shows that prevailing approaches perform 6% to 61% worse than our method while also being more time inefficient. Ablation studies show how the parts of the new approach contribute to the overall metrics.
title Ordered Semantically Diverse Sampling for Textual Data
topic Computation and Language
Information Retrieval
url https://arxiv.org/abs/2503.10698