🔢 mathematics

Roughness and entropy measures of a soft set

This paper introduces and systematically investigates two novel roughness measures and six entropy measures for soft sets, preserving Molodtsov's foundational principles while demonstrating their theoretical distinctiveness from classical rough set theory.

Original authors: Santanu Acharjee, Sankar K. Pal

Published 2026-05-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Santanu Acharjee, Sankar K. Pal

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to organize a messy room full of objects. Some objects are clearly a "chair," others are clearly a "table," but some are blurry, overlapping, or half-hidden. How do you decide what belongs where?

This paper tackles that exact problem, but in the world of mathematics and computer science. The authors, Santanu Acharjee and Sankar K. Pal, are trying to improve how computers handle uncertainty and imprecision.

Here is a simple breakdown of what they did, using everyday analogies.

1. The Problem: The "Blurry Photo" Issue

For decades, mathematicians have used a tool called Rough Set Theory to deal with blurry data. Imagine you have a photo of a crowd. If you try to draw a line around "the people wearing red," and two people are standing right next to each other, the old math tools get confused. They can't easily tell where one person ends and the other begins if they overlap. The old tools treat the whole messy blob as one big, undefined shape.

The authors say: "We need a better way to look at these overlaps."

2. The Solution: The "Attribute Tag" System

To fix this, they use a concept called Soft Set Theory. Think of this as giving every object in your room a set of tags or attributes.

Instead of just saying "This is a chair," you say "This is a chair that is wooden, brown, and heavy."
In their math, these tags are called attributes.

The authors realized that if you look at an object through the lens of its specific tags (attributes), you can see overlaps much more clearly. It's like looking at a Venn diagram where every circle is labeled with a specific feature.

3. The New Tools: Measuring "Fuzziness"

The paper introduces two main new tools to measure how "fuzzy" or "rough" a situation is:

A. Roughness Measures (The "How Messy Is It?" Ruler)
They created two new ways to measure the messiness of a group of objects.

The Analogy: Imagine you are trying to fit a square peg into a round hole.
- Old Method: You just say, "It doesn't fit perfectly."
- New Method: They created a ruler that measures exactly how much of the peg is sticking out and how much is inside, even if the peg is squished or overlapping with another peg.
They proved that their new rulers work better when objects overlap, which is a common problem in things like image processing (identifying objects in a photo).

B. Entropy Measures (The "Confusion" Meter)
In science, "entropy" usually means disorder or confusion. The authors created six different meters to measure how confused a computer is when looking at a soft set.

The Analogy: Imagine you are trying to guess a secret code.
- If the code is obvious, your "confusion meter" is low.
- If the code is a jumbled mess, the meter is high.
They designed six different types of meters:
- Some use logarithms (like a standard ruler).
- Some use exponents (like a magnifying glass that zooms in on the mess).
- Some are fixed (always work the same way).
- Some are adjustable (you can turn a dial to make them more or less sensitive).

4. The Big Discovery: One Size Does Not Fit All

After testing all six "confusion meters," the authors found something important: There is no single "best" meter.

Meter #1 (The Simple One): It's fast and easy to use, like a quick glance. But, it might miss some details if you are trying to analyze a complex picture where the background and the object are mixed together.
Meter #3 (The Detailed One): It's a bit more complex, but it follows stricter rules (axioms) and gives a more accurate picture of the confusion, especially when you need to look at the object and its background at the same time.

The Takeaway: Just like you wouldn't use a hammer to fix a watch, you shouldn't use the same math tool for every problem. If you need speed, use the simple meter. If you need high precision for complex overlaps, use the detailed one.

5. Why This Matters (According to the Paper)

The authors claim that their new way of measuring "roughness" and "confusion" is better than the old methods because:

It respects the original rules of Soft Set Theory (which some previous attempts ignored).
It can handle overlapping objects much better.
It offers a toolkit of different tools so experts can choose the right one for their specific job.

They specifically mention that this could be very useful for pattern recognition (teaching computers to recognize shapes) and image processing (cleaning up or analyzing photos), especially when objects in the photo are overlapping or blurry. They also hint at its potential use in space exploration (analyzing images of planets), but they keep the focus on the mathematical framework itself.

Summary

Think of this paper as the invention of a new, multi-tool Swiss Army knife for handling messy data.

The old knife (Rough Set Theory) struggled when two things overlapped.
The new knife (Soft Rough Sets) uses "tags" to separate the overlaps.
The authors added six different blades (Entropy measures) to this knife, each designed for a specific type of mess.
They proved that while some blades are faster, others are sharper, and the best choice depends entirely on the job you are trying to do.

Technical Summary: Roughness and Entropy Measures of a Soft Set

Problem Statement
The paper addresses the limitations of existing hybrid models that combine soft set theory and rough set theory. While rough set theory (Pawlak) effectively handles uncertainty through approximation spaces and indiscernibility relations, it struggles in scenarios involving overlapping objects, such as in pattern recognition and image processing. Specifically, classical rough set definitions of lower and upper approximations cannot clearly distinguish the boundaries of two overlapping objects. Furthermore, previous attempts to integrate soft sets with rough sets (e.g., by Feng et al., Ali, Shabir et al.) often deviate from the foundational mathematical and philosophical principles established by Molodtsov for soft set theory. These deviations include the incorrect use of soft subsets and operations that do not align with Molodtsov's original definitions, particularly regarding the treatment of attribute sets and the non-existence of "soft subsets" in the traditional sense. Consequently, there is a lack of a rigorous theoretical framework for defining roughness and entropy measures for soft sets that strictly adheres to Molodtsov's axioms while effectively handling overlapping approximations.

Methodology
The authors develop a novel framework for "soft rough sets" by strictly adhering to Molodtsov's foundational principles. The methodology involves:

Redefining Approximations: Instead of using subsets of the universal set directly, the authors define lower and upper soft approximations based on the association of attributes with approximate subsets.
- Lower Soft Approximation ( $L(S, A)$ ): Defined as the collection of pairs $(a_i, \cup B_j)$ where $B_j$ are equivalence classes contained within the attribute subset $S(a_i)$ .
- Upper Soft Approximation ( $U(S, A)$ ): Defined as the collection of pairs $(a_i, \cup C_k)$ where $C_k$ are equivalence classes having a non-empty intersection with $S(a_i)$ .
- Boundary: The boundary is defined as the difference between the upper and lower soft approximations for each attribute.
Roughness Measures: Two distinct roughness measures are introduced:
- Pawlak-style ( $\theta^P_R$ ): Based on the ratio of the cardinality of the lower approximation to the upper approximation, adapted for the sum over all attributes.
- Yao-inspired ( $\theta^Y_R$ ): A new measure designed to satisfy a specific set of axioms for accuracy (including conditions for exact sets, empty sets, and monotonicity with respect to granulation). This measure incorporates both the lower approximation and the complement's upper approximation to ensure symmetry between an object and its background.
Entropy Measures: Six entropy measures are proposed to quantify the incompleteness of knowledge associated with a soft set:
- Three logarithmic measures (based on Shannon's entropy): Two based on the Pawlak roughness measure (involving the object and its complement) and one based on the Yao-inspired roughness measure.
- Three exponential measures: Two based on the Pawlak roughness and one based on the Yao-inspired roughness.
- These measures are analyzed for properties such as maximum value, continuity, symmetry, and concavity.
Theoretical Analysis: The paper rigorously proves properties regarding internal and external approximations, the behavior of these measures under different equivalence relations (granularity), and the relationship between the proposed measures and their complements.

Key Contributions

Novel Roughness Framework: The paper introduces a concept of soft rough sets that preserves Molodtsov's original definitions, specifically avoiding the concept of "soft subsets" and utilizing internal/external approximations correctly. This allows for the distinct definition of lower and upper approximations even when objects overlap.
Two Roughness Measures: The proposal of a Pawlak-style measure and a new Yao-inspired measure that satisfies specific axiomatic requirements, offering a more robust way to assess accuracy in soft sets.
Six Entropy Measures: The systematic definition and analysis of six entropy measures (three logarithmic, three exponential) derived from the proposed roughness measures.
Axiomatic Validation: The demonstration that the proposed Yao-inspired accuracy measure satisfies a set of axioms that the Pawlak-style measure does not, particularly regarding the symmetry between a set and its complement.
Comparative Analysis: A detailed comparison showing that the proposed framework resolves the issue of overlapping regions that classical rough set theory cannot handle, as illustrated through theoretical examples and visual representations.

Results

Theoretical Properties: The authors prove that if one soft set internally approximates another, their lower and upper soft approximations maintain this relationship. Conversely, for external approximations, the relationship holds for lower approximations but not necessarily for upper approximations.
Granularity Impact: The study confirms that as the granulation of the universal set becomes finer (equivalence classes become smaller), the accuracy of the soft set increases, and the roughness decreases.
Entropy Behavior:
- The logarithmic entropy measures are shown to be concave and continuous.
- The exponential entropy measures exhibit specific monotonic behaviors dependent on the parameter $\beta$ .
- The measure $Ent^1_P_e$ (Pawlak-based, two-variable) is computationally efficient but lacks axiomatic consistency regarding the complement.
- The measure $Ent^3_P_e$ (Yao-based, one-variable) is identified as more theoretically consistent and suitable for general-purpose modeling, particularly when analyzing objects and backgrounds simultaneously.
Handling Overlaps: The proposed definitions successfully distinguish the boundaries of overlapping regions (e.g., red, blue, and green regions in an image) by treating them as distinct attribute-based approximations, a feat not possible with standard Pawlak approximations.

Significance
The paper claims that its primary significance lies in providing a mathematically sound and philosophically consistent framework for analyzing uncertainty in soft sets, specifically addressing the limitations of existing hybrid models. By strictly preserving Molodtsov's principles, the proposed framework offers a novel approach to roughness characterization that is distinct from classical rough set theory.

The authors highlight that this framework is particularly valuable for applications involving overlapping objects, such as pattern recognition, image processing, and AI vision. They suggest that the ability to distinctly define lower and upper approximations for overlapping regions makes the methodology a potential complementary tool to rough set theory in complex domains like space exploration (e.g., analyzing exoplanetary surfaces). The paper concludes that while no single entropy measure is universally optimal, the proposed family of measures offers a flexible toolkit for uncertainty quantification, with the choice of measure depending on the specific requirements of the application and the need for axiomatic consistency versus computational simplicity.