Real-world decision support in hospitality using pessimistic multi-granulation roughness of cubic intuitionistic fuzzy soft sets Authors: Asma Bibi, Shahida Bashir, Muhammad Shabir, Rabia Mazhar, Muhammad I Syam, Rizwan Gul Published: 2026 ABSTRACT AbstractTraditional fuzzy and rough set models usually have a hard time in providing sufficient ambiguity, hesitancy, and partiality to information that is common in actual hospitality decision-making. In order to overcome these problems, this paper suggests a new decision-support model that is built upon pessimistic multi-granulation rough sets combined with cubic intuitionistic fuzzy soft relation. The model proposed is a unification of interval-valued membership, intuitionistic non-membership and soft binary relations in a structure of pessimistic rough approximations. The framework characterizes multi-dimensional uncertainty of complex hospitality appraisals by modelling membership, non-membership, hesitation and boundary regions together in a variety of granulations. Soft binary relations defined in terms of foresets and aftersets are used to form lower and upper pessimistic multi-granulation approximations of internal cubic intuitionistic fuzzy sets. Approximation operators are formally defined in two pairs and the algebraic property of the operators is explored. In addition, measures of similarity among internal cubic intuitionistic fuzzy sets on soft relational structures are created to facilitate the robust alternative comparison. On the grounds of these theoretical bases, a systematic decision-making scheme that is composed of two structured algorithms is set up. The relevance of the suggested framework can be proved by a practical hotel selection case study with the usage of several factors and professional evaluations. Comparative and sensitivity analysis demonstrates that the model proposed gives consistent rankings and minimizes the effects of boundary ambiguity in different levels of uncertainty, which can be used as a trusted and risk-conscious decision support device in hospitality management.Keywords: Cubic fuzzy set, Fuzzy set, Rough set, Intuitionistic fuzzy set, HospitalitySubject terms: Engineering, Mathematics and computing Introduction Complex situations entail uncertain, imprecise and incomplete information in making decisions. Fuzzy sets (FSs), intuitionistic fuzzy sets (IFSs), and their types are advanced technologies that have been developed to address intricate situations in which classical methods of mathematics have been proven to be inadequate. Among these methods, internal cubic intuitionistic fuzzy sets (ICIFSs) are bound to be mentioned as one of the powerful tools, and it is comprised of interval-valued membership and a clear statement about hesitation. This bi-structured system provides a more elaborate format on which to state real-world problems with ambiguous and comprehensive information. The paper is based on the methodology that the study of Zeeshan et al.1 uses in the context of FSs and enriches it with the assistance of the fundamental principles of Ali et al.2 and Bashir et al.3. Compared to the case with conventional FSs, cubic fuzzy sets (CFSs) demonstrate a higher ability to embrace ambiguous data and discover their broader use in such areas as pattern recognition and in healthcare diagnostics. To this concept of soft relations (SRs), therefore, there is a decision support model proposed in the current work. There is a practical example given in order to demonstrate the applicability of the framework. The cubic fuzzy rough soft set (CFRS) model introduces the concepts of rough set (RSs), CFSs and soft set (SSs) and thus enables the ability to work with more imprecise and uncertain data more robustly. In the ever-evolving and highly dynamic scenarios where quick decisions need to be made, alternative strategies continue to get better in terms of handling uncertainty. It should be mentioned that FSs, SSs and RSs have become potent instruments one can use to capture and analyze vagueness, inconsistency and imprecision of data. In such fast-paced and dynamic situation where urgent decisions have to be made, several methods come to mind when approaching uncertainty in a more pragmatic manner. FSs, SSs, and RSs are excellent techniques of modeling discrepancies of vagueness, inconsistency and imprecision in data. The two frameworks have their benefits and are used in the management of uncertainty. These are applied in numerous applications in diverse fields, including computer science, decision sciences, and other practical applications. In practice, most reasoning models fail to achieve satisfactory results because of the abundance of incomplete or ambiguous data that is typical in the real-life case. The gap between the outdated mathematical strategies and the need to address uncertainty has been narrowed down over the past few years. The model of uncertainty representation differs depending on the discipline, though recent studies focus on enhancing the existing theoretical approaches and render them more efficient4. The concept of FSs was first presented by Zadeh5 as an extension of classical set theory with which to handle uncertainty more effectively. FSs provide a mathematically defined means of modeling the vague linguistic terms, and this is why they can be used in several fields. Their major limitation though is that they can merely fall back on the values of membership and this is normally not sufficient to represent the ambiguity or inaccurateness of uncertain or imprecise information in the real world. To overcome this shortcoming, Atanassov6,7 proposed the IFS as a combination of the FS with additions of two new parameters, non-membership and degrees of hesitation, in addition to membership. The three components of membership, non-membership, and hesitation are introduced in the structure to provide a more specific and diverse concept of depicting uncertainty. Due to this excellent structure, IFS has been applied widely in fields like machine learning, healthcare diagnostics, decision-making systems such as career and candidate selection, and pattern analysis and recognition8,9. Another approach to uncertainty that was proposed by Pawlak10 is called the RS theory, and it relies on approximation methods. It is a model of data that characterizes it through lower and upper approximations of equivalence relations11 enabling the identification of exact and vague elements in a dataset. Even though this is a useful approach in most areas, its use tends to be limited to practical application due to its dependence on strict equivalence relations. The RSs help in decision making in circumstances where ambiguity has been noticed e.g. in the sphere of medical diagnostics. They can be used as an example in the diagnosis of abdominal pain in children, where acquisition of correct classifications is a major issue. Pawlak’s rough set theory depends on a single equivalence relation. For dealing with multi challenging problems, Qian et al. 12 introduced a multigranulation rough set theory to approximate the subset of a universe with respect to multi equivalence relations instead of single-equivalence relations. In the struggle to eliminate the drawbacks of both FS and RS theories, Molodtsov13 developed the SS theory. This model was a parameterized approach to handling uncertainty that lacked much of the restrictions encountered in the previous models. SSs possess a flexible and mathematically sound framework and enough parameters to put across the information of uncertainty. It is in this light that more research continues to be undertaken in order to further build the SS theory by including a number of operations, which may assist in rendering it more relevant in more complicated decision-making circumstances. In particular Ali et al.14 contributed to such development by suggesting the generalization of operations with the position of the weaker parameter requirements that make the model more flexible and broaden the range of its application. The application of SSs in practice is repeatedly demonstrated in other researches and has been found in many applications present in the literature15,16. Numerous developments have been made to generalize the basic SS theory, to improve its performance in the case of uncertain information handling. These are intuitionistic fuzzy soft sets (IFSSs), fuzzy bipolar SSs, objective SSs and probabilistic SSs, which provide different means to address different aspects of uncertainty. In addition to this, it suggests a fuzzy RS method, based on coverings, and elaborates t-norms with implication of fuzzy logic to generate superior approximation methods. Similarly, a time-aggregated model of an infinite sequence of IFSs was suggested by Alcantud et al.17. They represent and study intuitionistic fuzzy data over time and are more dynamic and time-sensitive by employing historical scores and accuracy measures. In an even more general form fuzzy binary relations are generalizations of classical binary relations and are characterized as parameterized families over conventional relational models. In the traditional way of RS method, approximations are normally obtained following a single binary relation. This can however be bypassed with the concept of multiplicity of binary relations through the use of the soft binary relations (SℬRs). Feng et al.15 provided the concept of an soft relation over a universal set, which was the foundation of the application of relational operations in the SS paradigm. The works by Babitha and Sunil18 continue the theme of features and applications of SS relations and, thus, enhance the theoretical context and the real-life setting of the given sphere. Many literature works have been developed based on the original RS framework as proposed by Pawlak by using diverse relational structures, such as tolerance relations, covering-based relations, supremacy and community relations, and parallel relations. These improvements are also aimed at making RS more flexible and applicable to the uncertainty management of more complex information systems. The paper is focused on the construction of a framework to analyze the roughness of ICIFSs regarding SRs and the ways of their applications to a decision-making scenario. The proposed solution has a solid foundation of the case studies, which demonstrate its ability to model the multidimensional interdependence of the decision parameters and produce precise and consistent output. The data-driven uncertainty modeling, multi-criteria decision-making, and computational intelligence are also given value through the methodology. Anwer et al.19 worked on optimistic multi-granulation roughness with soft relations in the year (2021) thus; has more flexibility over uncertainty. Then, Anwar et al.20 designed an pessimistic multi-granulation model of IFSs over soft relations and emphasizes on conservative evaluations that are particularly relevant in complicated decision making scenarios. Jun et al. worked on cubic interval-valued IFS 21. Collectively, these works provide an opportunity to relate the benefits of multi-granulation of the RS theory to the ICIFSs where the levels of membership, non-membership, and hesitation will be introduced, and, thus, enable us to make more accurate approximations of the boundaries and more reliable decisions. Figure 1 shows the structure of the paper, and Table 18 presents the abbreviations used in the paper. Outline of the article. Table 18The list of acronyms.AcronymsRepresentationFSsFuzzy setsCFSCubic fuzzy setIFSsIntuitionistic fuzzy setCFRSCubic fuzzy rough setRSRough setCFSSCubic fuzzy soft setSSsSoft setsIFSSsIntuitionistic fuzzy soft setsICIFSInternal cubic intuitionistic fuzzy setMGRSMulti-granulation rough setPMGRSPessimistic multi-granulation rough setSℬRSoft binary relationMGGSASMulti-granulation generalized soft approximation set The list of acronyms. Recent advances and research gaps Recent development of decision-support methodologies proves that hybrid structures such that FSs, IFSs, RSs and SSs can enhance the quality of decisions made in a complex and uncertain environment significantly. Such hybrid models have been used more and more in hospitality management to solve multi-criteria evaluation problems like hotel ranking, quality of the service and analysis of customer preference. From these methods, it is emphasized that the parameterized uncertainty modeling should be used in conjunction with the structured approximation mechanisms. Even though the current research has been able to examine the incorporation of interval-valued knowledge, interval-valued approximation of hesitation, and rough approximations, they have seldom examined the incorporation of multi-granulation rough reasoning, cubic intuitionistic fuzziness, and soft relational dependencies within the same coherent framework. Specifically, modern models of decision-support in such a way that current models in the field of hospitality lack: a system to express both simultaneously interval based membership and intuitionistic non-membership; a system to express both simultaneously interval based membership and intuitionistic non-membership; a systematic means to model relational interactions of a parameter-dependent kind; and a systematic means to model relational interactions of a parameter-dependent kind; and a risk-averse multi-granulation strategy which fits in risk-conscious and conservative decision making. a risk-averse multi-granulation strategy which fits in risk-conscious and conservative decision making. Consequently, the entire range of uncertainty that also includes membership, non-membership, hesitation, relational dependency and boundary ambiguity is yet to be adequately reflected in the existing hospitality decision systems in the real world. This paper intends to solve these limitations by suggesting a Real-World Decision Support system in Hospitality using Pessimistic Multi-Granulation Rough Sets with Cubic Intuitionistic Fuzzy Soft Relations. In the proposed model, it is combined: ICIFSs to reflect interval-valued membership as well as intuitionistic non-membership so that to model the uncertainty precisely and at an interval. ICIFSs to reflect interval-valued membership as well as intuitionistic non-membership so that to model the uncertainty precisely and at an interval. SℬRs to model parameter-dependent interrelationships between criteria of hospitality evaluation. SℬRs to model parameter-dependent interrelationships between criteria of hospitality evaluation. Multi-granulation rough set (MGRS) approximations that are pessimistic to build lower and upper boundaries in decision making in a multi-relational granulation. Multi-granulation rough set (MGRS) approximations that are pessimistic to build lower and upper boundaries in decision making in a multi-relational granulation. An integrated framework that both characterizes membership, non-membership, hesitation and boundary areas in a risk-constrained decision-support paradigm. An integrated framework that both characterizes membership, non-membership, hesitation and boundary areas in a risk-constrained decision-support paradigm. Formal definitions of two new pessimistic multi-granulation rough approximation operators are given and structured decision-making algorithms are created. The relevance of the framework is illustrated by one of the real-world hospitality cases based on hotel choice under various evaluation criteria and professional beliefs. The proposed framework offers an effective and coherent instrument of addressing complex and real-world cases of decisions in hospitality management by utilizing relational granularity and cubic intuitionistic uncertainty with a pessimistic strategy. It also helps in the development of progress in uncertainty modeling, multi-criteria decision analysis and computational intelligence through provision of a theoretically based though practically applicable decision-support system. Contributions The following are the notable contributions of this paper: In this paper, we offer an original decision-support structure of hospitality management, which is a pessimistic multi-granulation rough sets and cubic intuitionistic fuzzy soft relations. In this paper, we offer an original decision-support structure of hospitality management, which is a pessimistic multi-granulation rough sets and cubic intuitionistic fuzzy soft relations. We define new pessimistic multi-granulation rough approximation operators of internal cubic intuitionistic fuzzy set and investigate its basic properties under soft relational structures. We define new pessimistic multi-granulation rough approximation operators of internal cubic intuitionistic fuzzy set and investigate its basic properties under soft relational structures. We also use soft binary relation to represent parameter dependent inter-relationships between hospitality evaluation criteria to increase realism in making multi-criteria decisions. We also use soft binary relation to represent parameter dependent inter-relationships between hospitality evaluation criteria to increase realism in making multi-criteria decisions. We come up with systematic decision-making algorithms which syntactically convert theoretical approximations into practical ranking and selection processes. We come up with systematic decision-making algorithms which syntactically convert theoretical approximations into practical ranking and selection processes. We test the proposed framework using a real-life case study of hotel selection with comparative and sensitivity analysis and prove that it is sound in interval and relational uncertainty. We test the proposed framework using a real-life case study of hotel selection with comparative and sensitivity analysis and prove that it is sound in interval and relational uncertainty. Organization of Paper FSs, SSs, RSs, IFSs, CIFSs, interval-valued fuzzy sets, interval-valued intuitionistic fuzzy sets, CFSs, ICFSs, and soft binary relations are among the fundamental theoretical ideas that are introduced in Section 2. In Section 3, we discuss how to approximate ICIFS. In Section 4, two SℬRs over dual universes are used to construct the notion of pessimistic multi-granulation roughness in the context of ICIFSs. Through illustrated examples, the algebraic features of this rough set model are examined in Section 5. A decision making algorithm based on the suggested framework is described in Section 6 along with a real world case study that illustrates its efficacy. To illustrate the benefits of the suggested method, Section 7 contrasts it with current theoretical models. Section 8 provides a summary of the research’s primary conclusions, while Section 9 addresses the study’s wider limitations. Preliminaries This area of the paper introduces several fundamental concepts related to binary relations: RS, MGRS, SS, CIFS and ICIFS, etc. We begin with the concept of a soft set (SS), which provides a flexible mathematical framework for representing objects under varying parameters, forming a basis for modeling uncertainty through parameterized structures. Definition 2.1 12Let k ⊆ L where L is a set of parameters. A pair (μ ,k) is said to be a SS over χ if μ is a mapping given by We also denote (μ ,k) by Definition 2.2 A binary relation R from is a subset of and a subset of is called a binary relation on χ. If R is a binary relation on χ, then R is said to be reflexive if symmetric if (u,v) ∈ R implies that and transitive if implies that If a binary relation is reflexive, symmetric and transitive, then it is called an equivalence relation. Definition 2.3 Let X ⊆ Y where Y is a set of parameters. A pair (η ,X) is called a SℬR over χ if (η ,X) is a SS over We also denote (η ,X) by In other words, a SℬR over χ is a parameterized family of binary relations on χ, that is, for each parameter we have a binary relation η (e) on χ. For X ⊆ Y, define SℬR over χ}. Definition 2.4 10It will be assumed that α is an equivalence relation on the universal set χ . The lower and upper approximations of P with respect to “α” for are given by while is the equivalence class of with respect to α . Boundary of is the set In the case, if then P is a rough estimate relative to “α”. Definition 2.5 5A set of ordered pairs representing a FS T defined over χ can be obtained as where shows the degree of membership of the element Definition 2.6 21An interval-value fuzzy set T on a universe χ is a mapping such that where Int ([0,1]) stands for the set of all closed subintervals of [0,1], the set of all interval-value fuzzy set on χ is denoted by Suppose that for all is called the degree of membership of an element where In certain situations, assigning an exact membership value becomes difficult because the available information may be limited or inconsistent. To address this issue, IVFS and IVIFS extend their classical models by using interval-valued memberships instead of single fixed values, allowing uncertainty to be represented in a more realistic manner. Definition 2.7 6Let χ be a non-empty universal set. An IFS T in the universe χ is represented as an entity of the form where and satisfy the condition for all The values and are called degree of membership and degree of non-membership of to T, respectively. The number is called the degree of hesitance of to T. Definition 2.8 21An interval value intuitionistic fuzzy set T defined over χ is given as with The pair is called an interval-valued intuitionistic fuzzy numbers. Definition 2.9 1Let (λ ,A) be a SℬR from and be IFS in Then, the lower approximation and the upper approximation of of with respect to aftersets are defined as follows: where and is called the afterset of for Definition 2.10 21A CFS is characterized by an ordered pair Where respectively denote the interval-value fuzzy set and fuzzy set. We shall denote this by To combine interval-valued and intuitionistic perspectives, the CIFS model and its internal form, ICIFS, were developed. These frameworks account for interval uncertainty, hesitation, and intuitionistic characteristics at the same time, creating a multidimensional structure that is particularly suitable for decision-making situations involving overlapping and incomplete information. Definition 2.11 21A cubic intuitionistic fuzzy set is said to be ICIFS if and for all Definition 2.12 In order to compare two or more ICIFSs, we have defined some metric methods in terms of score and accuracy functions for ICIFS which are given as follows: Roughness of internal cubic intuitionistic fuzzy set and its application in decision making This section of the study discusses how to approximate an ICIFS of and an ICIFS of applying aftersets and foresets of SℬR, respectively. By doing this, we can produce two internal cubic intuitionistic fuzzy soft sets in We also look at some of these approximation characteristics. Definition 3.1 22Let be a ICIFS in and be a SℬR from to , where are two interval-valued fuzzy sets and is an IFS. The relation between interval-valued fuzzy set and IFS is and We define pessimistic lower approximation and the upper approximation of with respect to foresets as follows: For all and are called the foresets for all and e ∈ S. Indeed and are two internal cubic intuitionistic fuzzy SSs over Definition 3.2 22Let be an ICIFS in and be a SℬR from to where are two interval-valued fuzzy sets and is an IFS. The relation between interval-valued fuzzy set and IFS is and We define lower approximation and upper approximation of with respect to aftersets as follows: For all and are called the foresets for and e ∈ S. Indeed and are two internal cubic intuitionistic fuzzy SSs over Theorem 3.3 22Let be a SℬR from to that’s, For any ICIFSs, and of following are true: Theorem 3.4 22Let be a SℬR from to that’s, For any ICIFS, and of following are true: Pessimistic multi-granulation roughness of cubic intuitionistic fuzzy sets and its applications in decision making Using foresets and aftersets formed from SℬRs, this component of the study focuses on approximating an ICIFS and an internal cubic fuzzy set. Specifically, we use SℬRs constructed across dual universes to investigate the pessimistic multi-granulation roughness of an ICIFS. The lower and upper approximations under the pessimistic framework are represented by two different ICIFSs that are produced by this method. In order to gain a better understanding of the structure and behavior of internal cubic intuitionistic fuzzy soft sets in this granular rough set environment, we also look into a number of important properties of these approximations. Definition 4.1 Let be two SℬRs from and be a ICIFS in where are two interval-valued fuzzy sets and is IFS. We define pessimistic lower approximation and the pessimistic upper approximation of with respect to aftersets as follows: For all are called the aftersets of for and e ∈ S. Indeed and are two internal cubic intuitionistic fuzzy SSs over Definition 4.2 Let be two SℬRs from and be a ICIFS in where are two interval-valued fuzzy sets and is IFS. We define pessimistic lower approximation and upper approximation of with respect to foresets as follows: For all are called the foresets of Indeed and are two internal cubic intuitionistic fuzzy SSs over Example 4.3 Let and are two SℬRs from defined by, and Next, the aftersets and foresets are computed as follows: and and Now define ICIFS , which indicates in Table 1. Table 1ICIFS .123[0.2,0.4][0.1,0.3][0.4,0.7][0.1,0.5][0.4,0.6][0.2,0.4]0.20.30.20.30.50.4 ICIFS . Therefore, pessimistic multi-granulation lower approximation and the pessimistic multi-granulation upper approximation of ICIFS concerning aftersets are according to Definition 4.1, are given as: concerning foresets are given concerning foresets are given which are respectively presented in Tables 2 and 3. Table 2Pessimistic multi-granulation lower approximation of ICIFS .efgh Pessimistic multi-granulation lower approximation of ICIFS . Table 3Pessimistic multi-granulation upper approximation of ICIFS .efgh Pessimistic multi-granulation upper approximation of ICIFS . Now define which indicates the ICIFS displayed in Table 4. Table 4ICIFS .efgh[0.2,0.5][0.3,0.4][0.5,0.7][0.3,0.6][0.3,0.4][0.4,0.6][0.3,0.4][0.5,0.6]0.40.20.30.40.50.60.40.5 ICIFS . Therefore, in the light of Definition 4.2, the pessimistic multi-granulation lower approximation and the pessimistic multi-granulation upper approximation of ICIFS concerning foresets are given as: , , which are respectively presented in Table 5 and 6. Table 5Pessimistic multi-granulation lower approximation of ICIFS .123 Pessimistic multi-granulation lower approximation of ICIFS . Table 6Pessimistic multi-granulation lower approximation of ICIFS .123 Pessimistic multi-granulation lower approximation of ICIFS . The following theorem provides a consistent framework for comparing and approximating uncertain information between two related domains through aftersets. Theorem 4.4 Let and be two SℬRs from , that is, and and ICIFS is Then following are true: Proof Let . Then, .Similarly, Let . Then,. Also, Let .Then, ..Let .Then, .. Hence, Let . Then, . Similarly, Let . Then,. Also, Let .Then, . . Let .Then, . . Hence, Let . Then, . Similarly, Let . Then, . Also, Let Then,Let .Then, .. Hence, Let . Then, . Similarly, Let . Then, . Also, Let Then, Let .Then, . . Hence, Theorem 4.5 Let and be two SℬRs from that is, and and ICIFS is Then the following are true: Proof This proof follows a similar approach to the proof of a Theorem 4.4. Theorem 4.6 Let be two SℬRs from that is, and are ICIFSs. Then the following hold: Proof Then for, Hence, Then for, Hence, Then, Thus,So, Then, Thus, So, Then, Thus,Hence, Then, Thus, Hence, Then, Thus,Hence, Then, Thus, Hence, Theorem 4.7 Let be two SℬRs from that is, and are ICIFSs. Then the following hold: Proof This proof follows a similar approach to the proof of a Theorem 4.6. Theorem 4.8 Let be two SℬRs from and are ICIFSs of . Then the following hold: Proof Thus,And, .Hence, Thus, And, . Hence, Thus,Hence, Thus, Hence, This shows that This shows that ,Similarly, Also, This shows that , Similarly, Also, This shows that Also, Similarly, Also, This shows that Also, Similarly, Also, This shows that Also, Similarly, Also, This shows that Also, Similarly, Also, This shows that Theorem 4.9 Let be two SℬRs from and that are and are ICIFSs of Then, the following hold: Proof This proof follows a similar approach to the proof of a Theorem 4.8. Roughness of an internal cubic intuitionistic fuzzy set across two universe by multi-soft relations Definition 5.1 Let be multi-granulation generalized soft approximation space (MGGSAS) from where be a finite collection of SℬRs from and be an ICIFS in where are two interval-valued fuzzy set and is intuitionistic fuzzy set, are related as We define pessimistic lower approximation and the upper approximation of with respect to aftersets are: For all, is known as the aftersets for and e ∈ A. Indeed are two internal cubic intuitionistic fuzzy SSs over Definition 5.2 Let be MGGSAS from where be a finite collection of SℬRs from and be an ICIFS in where are two interval-valued fuzzy set and is intuitionistic fuzzy set, are related as We define pessimistic lower approximation and the upper approximation of with respect to foresets are: For all, is known as the fore sets of for and e ∈ A. So, are two internal cubic intuitionistic fuzzy SSs over Theorem 5.3 Let be MGGSAS from and is ICIFS of Then below properties hold: Proof This proof takes the similar steps as in Theorem 4.4. Theorem 5.4 Let be MGGSAS from and is ICIFS of Then below properties hold: Proof This proof takes the similar steps as in Theorem 4.5. Theorem 5.5 Let be MGGSAS from and and are ICIFS of Then, Proof This proof takes the similar steps as in Theorem 4.6. Theorem 5.6 Let be MGGSAS from and and are ICIFS of Then, Proof This proof takes the similar steps as in Theorem 4.7. Theorem 5.7 Let be MGGSAS from and are ICIFS of Then, Proof This proof takes the similar steps as in Theorem 4.8. Theorem 5.8 Let be MGGSAS from and are ICIFS of Then, Proof This proof takes the similar steps as in Theorem 4.9. Application to decision-making problems Because real world situations are dynamic and complicated, making it difficult for individual specialists to come up with the best solutions, effective decision making is crucial in data analysis. To address challenges arising from imprecise information provided by multiple observers, Sun and Ma23 introduced a consensus driven algorithm for object recognition. Building on their work, Feng et al.24 refined the approach to overcome specific limitations. Shabir et al.19 advanced the PMGRS theory further by combining soft relations and crisp sets to produce a strong framework for decision making. Their model, which merges CIFSs with PMGRSs under a soft relational structure, supports flexible assessments and offers a holistic approach for navigating uncertainty in practical decision making contexts. The pessimistic multi-granulation roughness of CIFS model aids in medical diagnosis and sustainable development by assessing diagnostic alternatives and evaluating environmental effects, enhancing patient care and environmental management. This is crucial for ensuring the sustainable use of natural resources and addressing environmental issues. The pessimistic multi-granulation roughness of CIFS model provides financial guidance, risk consideration, and strategic decision making support for investors, identifying design solutions for improved performance and cost effectiveness. The pessimistic multi-granulation roughness of CIFS model aids in business and management by providing a framework for evaluating operational choices and strategies, enabling managers to make informed decisions and navigate challenging business environments. In this regard, Algorithms 1 and 2 are presented to illustrate the proposed model, while Fig. 2 depicts its flowchart. Visual representation of the proposed algorithm. Decision-making using aftersets. Decision making problem using foresets. The upcoming scenario demonstrates the step by step decision making process. The algorithm used to make better decisions as an owner is discussed in this example. Example 6.1 Suppose a traveler wants to book a hotel, and there are six options available that are divided in two groups: amenities {Wi-Fi, Gym, Restaurant} and location {Downtown, Beachside, AirPort}, the set represents the amenities, and represents the location. Let H = {set of hotels} = be the parameters. We have representing the relations of amenities and location available at hotel Let. , , Similarly, representing the relations of amenities and location available at hotel next month. Let The overview of this example is given in Fig. 3: General overview of Example 6.1 We have representing the relations of amenities and location available at hotel. Let Similarly, representing the relations of amenities and location available at hotel next month. Let . Then the aftersets and foresets are as follows: Now, define ICIFS in as given in Table 7. Table 7ICIFS .[0.2,0.4][0.3,0.5][0.4,0.6][0.1,0.3][0.2,0.4][0.4,0.5]0.20.30.40.30.40.5 ICIFS . Therefore, according to Definition 4.1, the pessimistic multi-granulation lower approximation and the pessimistic multi-granulation upper approximation concerning aftersets are given as: Which are respectively exhibited in Tables 8 and 9. Table 8Pessimistic multi-granulation lower approximation of ICIFS . Pessimistic multi-granulation lower approximation of ICIFS . Table 9The Pessimistic multi-granulation upper approximation of ICIFS .a1a2a3 The Pessimistic multi-granulation upper approximation of ICIFS . According to Definition 2.13, the score values are presented in Table 10: Table 10Values of score function.0.0250.00.00.00.00.00.0250.00.0250.00.0250.00.00.0250.00.0250.0250.050.0250.00.0250.0250.00.00.050.0250.075 Values of score function. From Table 10, we can see that is maximum. So, the traveler will choose the hotel The graphically representation is given in Fig. 4. Graphical representation of aftersets for Example 6.1 through bar graph. Now, define ICIFS of as given in Table 11. Table 11ICIFS .[0.2,0.6][0.4,0.6][0.2,0.7][0.1,0.4][0.3,0.4][0.3,0.4]0.20.40.30.30.40.4 ICIFS . Therefore, according to Definition 4.2, the pessimistic multi-granulation lower approximation and the pessimistic multi-granulation upper approximation of ICIFS about foresets are given as: which are respectively exhibited in Tables 12 and 13. Table 12Pessimistic multi-granulation lower approximation of ICIFS 2 β2. Pessimistic multi-granulation lower approximation of ICIFS 2 β2. Table 13Pessimistic multi-granulation upper approximation of ICIFS . Pessimistic multi-granulation upper approximation of ICIFS . Based on Definition 2.13, the score values are shown in Table 14: Table 14Values of score function.0.00.00.0250.0250.0250.0250.0250.0750.10.050.0750.0250.0250.0750.0250.150.1250.2750.050.050.0750.0750.0250.0250.1750.1250.3 Values of score function. From Table 14, we can see that is maximum. So, the traveler will choose the hotel Its . Graphically representation is given in Fig. 5. Graphical representation of foresets for Example 2 through bar graph. Comparison The pessimistic multi-granulation roughness of the CIFS model enhances conventional models by incorporating CIFS and pessimistic multi-granulation rough sets, allowing for partial memberships and capturing details and uncertainties in real world data Jun et al.21, Mubarak et al.25. Conventional and fuzzy set theories face challenges due to complicated uncertainties and multi-granular analysis requirements. RS theory addresses these issues but lacks depth in pessimistic decision-making Qian et al.12. The strengths of multi-granulation rough set theory and cubic intuitionistic fuzzy sets are combined in the concept of pessimistic multi-granulation roughness based on CIFSs. This integration aims to achieve more refined knowledge representation by incorporating multiple granulation levels and emphasizing cautious or pessimistic outcomes. CIFSs enhance the modeling of uncertainty by simultaneously considering membership, non-membership, and hesitation degrees within a cubic structure. This enriched representation makes the model particularly useful for complex decision making scenarios, such as risk analysis, medical diagnosis, and multi criteria evaluation. Quantitative analysis The roughness of CIFS under soft relations was compared by evaluating the cumulative score values of Anwer et al.1, Jun Ye21, and the proposed one. The proposed method received the biggest score among them, Jun Ye21 was second with Anwer et al.1 coming in the last. This result proves the high effectiveness of the suggested framework. The discrepancies in performance are obvious as higher separation margins are obtained by the suggested approach both in absolute and relative terms in comparison to the previous approaches. As percentages of improvements, these will underline the efficacy of the offered solution in minimizing uncertainties and assisting in making more precise decisions. Since the cumulative scores in upper left quadrant of cubic intuitionistic fuzzy soft relation set are larger, leading to the reduction in the regions of the boundaries and the consequent decrease in roughness, the proposed approach produces the strongest and the most credible approximations. Comparatively, the approach that Jun Ye uses is moderately rough whereas the approach used by Anwer et al. has the most pronounced boundary uncertainty. On the whole, these results prove that the suggested framework is the most effective in terms of ambiguity reduction and offers the best point of decision making as compared to the other two options. The proposed approaches were implemented using the given dataset and the results are provided in Tables 15 and 16. Additionally, the graphical depiction of the ranking results derived through different schemes is portrayed in Fig. 6a, b, respectively. Table 15Comparison based on various studies w.r.t aftersets.MethodsCumulative score valuesRankingAnwer et al.10.20.2- 0.1Jun Ye211.71.61.5Proposed0.0750.050.025 Comparison based on various studies w.r.t aftersets. Table 16Comparison based on various studies w.r.t foresets.MethodsCumulative score valuesRankingAnwer et al.10.4 0.1 - 0.1Jun Ye211.9 1.8 1.2Proposed0.3 0.275 0.1 Comparison based on various studies w.r.t foresets. (a) Comparison w.r.t. aftersets (b) Comparison w.r.t. foresets. Qualitative analysis The comparison of the cumulative score values derived using three different methods was used to evaluate the roughness of CIFSs under soft relations. The findings indicate that the most cumulative score was obtained with the proposed approach, the second method was in the middle of the findings, and the first method obtained the lowest. Such a ranking is a clear indication of the high competence of the proposed framework in terms of minimizing the uncertainty and delivering more reliable results. Absolutes and percentage changes in cumulative scores can be regarded as great quantitative indicators of the benefit of the proposed approach, and a bigger change indicates a better performance. Qualitatively, the larger score of the suggested approach means that it gives the least roughness as lower and upper approximations are smaller and the boundary region is smaller. The second approach shows medium performance as it is able to grasp the uncertainty better than the last-ranked approach, but the range of the boundary is broader. On the contrary, the former approach is as rough as it is the highest-scoring with a higher level of ambiguity. On the whole, the examination confirms that the suggested framework is most efficient in roughness reduction and provides the most precise decision support among others. Table 17 presents the qualitative analysis of this model to existing models. Table 17Characteristics comparison.Methods CharacteristicsHandle membership degreeIndeterminacy membership degreeFalsity membership degreeRoughness of an information systemInternal cubic setParamete-rizationsDual universesRS (Pawlak)10✗✗✗✓✗✗✗IFS (Atanassov)6✓✓✓✗✗✗✗CFS (Jun Ye)21✓✓✗✗✗✗✗MGRS (Qian et al.)12✗✗✗✓✗✗✓IVIF (Li26)✓✓✓✗✓✗✗Approximation of IFS (Anwer et al.)1✓✓✓✓✗✓✗PMGRS of IFS (Anwer et al.)20✓✓✓✓✗✓✓Proposed✓✓✓✓✓✓✓ Characteristics comparison. Runtime/scalability discussion To obtain some empirical insights into the runtime performance, computational experiments with datasets of different sizes were carried out. The results show that the computational time increases with the number of objects and parameters, although the increase is gradual and can be classified as polynomial, as supported by the theoretical analysis of the computational complexity. For example, when the number of objects increased from n to m, the computational time increased proportionally although it remained within the desired limit. Special cases of the proposed model Showing how the proposed framework reduces to existing classical models under specific assumptions. Single-granulation rough soft model: Single-granulation rough soft model: When the multi-granulation rough set structure is chosen to be pessimistic with one granulation, one will have a classical soft rough set model. Multi-granulation aggregation (P-union / P-intersection) is eliminated. Multi-granulation aggregation (P-union / P-intersection) is eliminated. It becomes one rough approximating system of relationships that are soft. It becomes one rough approximating system of relationships that are soft. The case is an example of standard rough decision-support with relational dependency but no multi-source granularity. (2) Optimistic multi-granulation case: Optimistic multi-granulation case: When the pessimistic aggregation approach is substituted by the optimistic one, the model becomes an optimistic multi-granulation rough set model. When approximations become lower, they become loose. When approximations become lower, they become loose. Decision boundaries expand. Decision boundaries expand. Appropriate when the opportunity is being pursued and not when the decision makers are risk-averse. Appropriate when the opportunity is being pursued and not when the decision makers are risk-averse. In such a way, the pessimistic model is the generalization of the optimistic multi-granulation rough methods. (3)Interval-valued intuitionistic fuzzy soft model: Interval-valued intuitionistic fuzzy soft model: When cubic structure conditions are loosened and only interval-valued intuitionistic components are kept, then the framework becomes an interval-valued intuitionistic fuzzy soft set model. Boundary rough approximations would disappear. Boundary rough approximations would disappear. There are no lower/ upper rough areas. There are no lower/ upper rough areas. Membership, non-membership and hesitation are all that are left. Membership, non-membership and hesitation are all that are left. This indicates that the model proposed is an extension of classical interval valued intuitionistic fuzzy decision systems. (4) Classical intuitionistic fuzzy rough model: Classical intuitionistic fuzzy rough model: If: Equivalence relations are used in place of soft binary relations. Equivalence relations are used in place of soft binary relations. Membership interval-valued collapses to actual values, then the model becomes an intuitionistic fuzzy rough set model. Membership interval-valued collapses to actual values, then the model becomes an intuitionistic fuzzy rough set model. This goes to show that the suggested structure is a generalization of the intuitionistic fuzzy rough sets that introduces relational softness and interval uncertainty. (5) Classical rough set model: Classical rough set model: If: The values of membership and non-membership are limited to crisp values {0,1}. The values of membership and non-membership are limited to crisp values {0,1}. Intuitionistic elements are cubic elements, and Intuitionistic elements are cubic elements, and Soft relations narrow down to equivalence relations, the model reduces to the original rough set model that was proposed by Pawlak. Soft relations narrow down to equivalence relations, the model reduces to the original rough set model that was proposed by Pawlak. In this way, the most restrictive special case is classical rough sets. (6) Classical fuzzy soft decision model. Classical fuzzy soft decision model. When the rough approximations are eliminated and only cubic intuitionistic fuzzy soft relations are kept, the system will be reduced to a fuzzy soft decision-support model without the analysis of the boundary regions. In this case: None of the lower/upper approximations, None of the lower/upper approximations, No boundary region modeling, No boundary region modeling, Also left are only fuzzy aggregation and scoring. Also left are only fuzzy aggregation and scoring. (7) Super crisp multi-criteria decision-making case: Super crisp multi-criteria decision-making case: If: Any fuzzy and rough uncertainty elements are eliminated, Any fuzzy and rough uncertainty elements are eliminated, Criteria appraisals are accurate numerical values, Criteria appraisals are accurate numerical values, the model is a normal deterministic Multi-Criteria Decision-Making framework. This is the most basic restrictive scenario of the suggested structure. Limitations Limitations in the established framework can be summarized as follows: Complexity and scalability of computations Complexity and scalability of computations Combining pessimistic multi-granulation rough sets with cubic intuitionistic fuzzy soft relations greatly adds to the complexity of the computation. This difficulty increases with the increase in the number of alternatives, criteria, relations as well as the granulation levels. Therefore, the model can be expected to have scalability problems in the application to large scale, real world, decision making problems, which may limit its efficiency and practical application. (2) Sensitivity to the choice of parameters and the model Sensitivity to the choice of parameters and the model Soft binary relations, granulation structures, and aggregation strategies are the key components of the framework, the success of which strongly relies on their proper choice. The wrong or inefficient selection of parameters may cause fluctuating rankings and unreliable decision results. Thus, these components need to be carefully calibrated and validated to be robust and consistent in results. (3) Subjectivity, consistency and interpretability concerns Subjectivity, consistency and interpretability concerns The model uses interval-based membership and non-membership functions, and expert-based relational judgments, which pre determinately inject subjectivity in the decision-making process. This may result in inconsistencies across evaluations. Moreover, cubic intuitionistic fuzzy logic, together with multi-granulation rough approximations and soft relational modeling provide a mathematically complex structure an interpretation and application of which might prove challenging to practitioners without special expertise. (4) Conserve nature and low generalizability Conserve nature and low generalizability Although the pessimistic multi-granulation approach is appropriate in risk-averse decision-making environments, it can be overly conservative, which may rule out other viable options that can be acceptable in opportunity-driven settings. Moreover, the model presupposes a fixed decision environment and has been confirmed only on a few case studies. More widespread empirical testing on a variety of hospitality applications is needed to determine its generalizability, flexibility and overall reliability. Conclusions and future research plans The proposed research has introduced a new decision-support framework of hospitality management that relies on the pessimistic multi-granulation rough sets that are combined with the cubic intuitionistic fuzzy soft relations. The given model is the systematic combination of interval-valued membership, intuitionistic non-membership, hedges of the degree of hesitation, soft dependencies of relations, and rough approximations of boundaries within one framework. The framework allows risk conscious and conservative decision making that is appropriate especially in the real world hospitality context, where uncertainty, interdependence of criteria and incomplete information limit the decision making process. Two new multi-granulation rough approximation operators that are pessimistic were formally introduced in relation to ICIFSs of soft relational settings and their essential properties were examined. Development of structured decision making algorithm was done to translate the theoretical framework into practical ranking and selection processes. The model was proven applicable in a real-life hospitality case study on the hotel selection accompanied by comparative and sensitivity analysis which revealed that the decision outcome remained stable and consistent in different degrees of uncertainty. Although it contributes to the theory and practice, there are still a few areas in which the research can be driven. First, computational optimization methods can be created to increase the level of scalability of large-scale hospitality datasets. Second, the model can be extended dynamically or with respect to time to facilitate real-time decision-making in changing service environments. Third, mechanism of hybrid aggregation that incorporate both pessimistic and optimistic approaches can offer flexible decision modalities that can be used in both risk-averse and opportunity-driven situations. Fourth, it could be integrated with machine learning or data-based methods of parameter estimation to minimize the use of purely expert-based inputs. Lastly, the generalizability and strength of the proposed framework would be further demonstrated by closer empirical validation in the various hospitality industries, including planning of tourism services, benchmarking of service quality, and review of customer satisfaction. Generally, the research contributes to uncertainty modelling and multi-criteria decision-making in the hospitality industry by developing a mathematically sound but pragmatically feasible decision-support framework that is capable of dealing with multi-dimensional uncertainty in real-life situations. Acronyms list is given in Table 18. Author contributions Conceptualization was performed by S.B. and M.S. Methodology was developed by S.B. and M.S. Sofware was provided by S.B. Validation was performed by I.S. Formal analysis was performed by S.B., R.M. and R.G. Investigation was performed by A.B. Resources were provided by I.S. Data curation was performed by A.B., M.S. and S.B. Original draft was written by S.B. Reviewing and editing were performed by S.B. and A.B. Visualization was performed by S.B. and R.M. Supervision was done by S.B. Project administration was performed by S.B. Funding acquisition was done by I.S. All authors have read and agreed to the published version of the manuscript. Funding The authors would like to express their gratitude to the United Arab Emirates University, Al-Ain, UAE, for providing financial support with Grant No. G00004702. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Ethical approval Since neither humans nor animals will be used in this study, ethics committee approval is not required. No personally identifying information is gathered or analyzed; all data used in the study is publically available. Every technique closely adheres to accepted ethical norms and principles. Declaration of generative AI and AI-assisted technologies The writers affirm that they did not create this article using artificial intelligence (AI) methods. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.