Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law

Conventional term-weighting techniques in information retrieval, such as term frequency-inverse document frequency and the probabilistic ranking framework, often place excessive emphasis on extreme term frequencies, which can distort document ranking. Inspired by Lenz’s Law in electromagn...

Full description

Saved in:
Bibliographic Details
Main Authors: Krishnan Batri, S. Lakshmi, R. Sowrirajan
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10925389/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849767589051367424
author Krishnan Batri
S. Lakshmi
R. Sowrirajan
author_facet Krishnan Batri
S. Lakshmi
R. Sowrirajan
author_sort Krishnan Batri
collection DOAJ
description Conventional term-weighting techniques in information retrieval, such as term frequency-inverse document frequency and the probabilistic ranking framework, often place excessive emphasis on extreme term frequencies, which can distort document ranking. Inspired by Lenz’s Law in electromagnetism, which naturally resists sudden fluctuations in magnetic flux, this work introduces a novel parabolic weighting mechanism. By applying a parabolic function to reduce the impact of excessively frequent terms while enhancing moderately occurring words, the proposed method achieves a balanced contribution of term frequency. The mathematical formulation promotes equilibrium in term weighting by integrating term frequency with inverse document frequency. Experiments conducted on benchmark datasets, including BBC News and 20 Newsgroups, demonstrate that the parabolic weighting mechanism outperforms traditional techniques, yielding measurable improvements in accuracy with classification models such as the support vector classifier (0.44 percent increase) and logistic regression (0.30 percent increase). Furthermore, statistical validation using Cohen’s effect size measure confirms the significance of performance improvements, while bootstrap analysis ensures the reliability of the observed gains. These results establish a strong foundation for future integration with neural information retrieval models and highlight the potential of the proposed approach in domain-specific applications, such as legal document analysis and biological literature search.
format Article
id doaj-art-bb5f3fc4a0254995b599b23e28f8bf91
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-bb5f3fc4a0254995b599b23e28f8bf912025-08-20T03:04:07ZengIEEEIEEE Access2169-35362025-01-0113543675438210.1109/ACCESS.2025.355096410925389Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s LawKrishnan Batri0https://orcid.org/0000-0003-4538-1305S. Lakshmi1https://orcid.org/0000-0003-2317-9552R. Sowrirajan2https://orcid.org/0000-0001-9556-5482School of Computer Science and Engineering, Jain (Deemed-to-be University), Bengaluru, Karnataka, IndiaDepartment of Electronics and Communication Engineering, Jain (Deemed-to-be University), Bengaluru, Karnataka, IndiaDepartment of Mathematics, Dr. N.G.P. Arts and Science College, Coimbatore, Tamil Nadu, IndiaConventional term-weighting techniques in information retrieval, such as term frequency-inverse document frequency and the probabilistic ranking framework, often place excessive emphasis on extreme term frequencies, which can distort document ranking. Inspired by Lenz’s Law in electromagnetism, which naturally resists sudden fluctuations in magnetic flux, this work introduces a novel parabolic weighting mechanism. By applying a parabolic function to reduce the impact of excessively frequent terms while enhancing moderately occurring words, the proposed method achieves a balanced contribution of term frequency. The mathematical formulation promotes equilibrium in term weighting by integrating term frequency with inverse document frequency. Experiments conducted on benchmark datasets, including BBC News and 20 Newsgroups, demonstrate that the parabolic weighting mechanism outperforms traditional techniques, yielding measurable improvements in accuracy with classification models such as the support vector classifier (0.44 percent increase) and logistic regression (0.30 percent increase). Furthermore, statistical validation using Cohen’s effect size measure confirms the significance of performance improvements, while bootstrap analysis ensures the reliability of the observed gains. These results establish a strong foundation for future integration with neural information retrieval models and highlight the potential of the proposed approach in domain-specific applications, such as legal document analysis and biological literature search.https://ieeexplore.ieee.org/document/10925389/Information retrievalterm weightingparabolic weightingLenz’s lawsemantic analysiscomputational linguistics
spellingShingle Krishnan Batri
S. Lakshmi
R. Sowrirajan
Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law
IEEE Access
Information retrieval
term weighting
parabolic weighting
Lenz’s law
semantic analysis
computational linguistics
title Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law
title_full Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law
title_fullStr Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law
title_full_unstemmed Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law
title_short Parabolic Weighting Mechanism in Information Retrieval: A Mathematical Analogy to Lenz’s Law
title_sort parabolic weighting mechanism in information retrieval a mathematical analogy to lenz x2019 s law
topic Information retrieval
term weighting
parabolic weighting
Lenz’s law
semantic analysis
computational linguistics
url https://ieeexplore.ieee.org/document/10925389/
work_keys_str_mv AT krishnanbatri parabolicweightingmechanismininformationretrievalamathematicalanalogytolenzx2019slaw
AT slakshmi parabolicweightingmechanismininformationretrievalamathematicalanalogytolenzx2019slaw
AT rsowrirajan parabolicweightingmechanismininformationretrievalamathematicalanalogytolenzx2019slaw