**1. Introduction and Overview**

In the current scientific debate, cities represent the contexts in which resources, capital, skills, and talents are concentrated and, at the same time, places where many challenges are concentrated regarding environmental (pollution, waste, climate change), economic (unemployment, social exclusion, well-being), political (instability in governance processes, lack of strategic planning), and cultural (training, creativity and innovation) dimensions. The city, therefore, allows to experiment the paradigm shift from linear economy to circular economy, in which to promote synergetic, fair and inclusive processes capable of activating new forms of urban productivity and social and economic innovation.

In particular, the European Commission identifies the cultural heritage of cities as the main driver of development and supports strategies in which it is considered one of the fundamental elements of possible transformations, which can be activated through mixed top down/bottom up approaches in the short and long term. From this perspective, adaptive reuse can play a decisive role not only in terms of increasing the life cycle of the heritage but also as an urban strategy capable of generating new, even divergent values (economic, aesthetic, cultural, educational, political), supporting innovative dynamics of local development and which underline the importance of preserving cultural heritage and the instrumental values for its livelihood [1] in a "circular" perspective [2]. In fact, if you look at cultural heritage as a potential element of an "self-sustainability" territorial system [3], which continually redefines itself and sustains and reproduces itself from within, it is possible to identify the re-use of cultural heritage as an essential contribution in the realization of a model of economy and of "circular" cities and territories, capable of eliminating the waste of environmental resources (soils, materials,

energy, etc.) and cultural resources (tangible and intangible heritage, knowledge, etc.), and capable of transforming "waste" into new economic, environmental, cultural, and social resources.

Cultural heritage, in fact, can be considered not only a legacy to be handed down to posterity but also a central resource for triggering processes of local and global development. At the moment, however, the conservation and enhancement of the heritage that has a cultural and landscape value represent a burden for the community. However, the inequality between the quantity of material and cultural heritage available and the increasingly limited availability of public resources for conservation have led to the total or partial abandonment of cultural heritage. Even where conservation measures are carried out, the potential impact of this public investment is countered by an absent or short-sighted management program.

Today, in the field of reusing historical heritage, scientific research has developed methods to assess the compatibility of new functions with the morphological and dimensional characteristics of existing buildings and with the objectives of urban development and revitalization [4–6]. In the planning and technical and economic feasibility study phase, the purpose of these methods is to determine the most compatible use. In the planning phase, the compatibility assessment allows decision-makers to identify new appropriate uses, compatible with the layout and the original meaning of buildings and historic places, a new use that will bring benefits such as the improvement of the building, market value, use, environmental, cultural and social value [5,7]. The desire to protect the identity of a historic building leads planners to estimate the quantity and amount of transformations required by each function in order to find the most compatible solution. However, the choice of the new function must also take into account the economic and social objectives and the effects of reuse on the urban environment.

The lack of public resources has, in fact, pushed many local administrations to favor new forms of enhancement of public buildings in able to promote the "adaptive reuse" of a historic building in order to preserve its social, historical, and cultural values but also to stimulate the experimentation of new circular businesses, financing and governance models in heritage conservation, creating synergies between multiple actors, and reducing the use of resources and regenerating values, capital, and knowledge

Decisions on the increasingly scarce allocation of public resources for the conservation or development of cultural heritage are based on a series of multiple and heterogeneous criteria, which are often in conflict, as well as on the preferences of various stakeholder. In this context, multi-criteria approaches provide an adequate theoretical and methodological framework to deal with the complexity that characterizes the adaptive reuse strategies of cultural heritage.

In the literature, many multi-criteria methods have been developed in recent decades [8] with applications to real cases (including examples of reuse) and in different sectors to support decision-making processes [9]. "Multi-criteria analyses" are very useful tools in the ex-ante evaluation phase, in which different options must be evaluated in order to determine and classify the alternatives. Therefore, some multi-criteria evaluation methods have been developed as "decision support systems": they do not automatically offer a choice, but rather provide help to decision-makers who face complex situations and problems characterized by a certain degree of uncertainty and information of different nature.

In order to evaluate the overall effectiveness of the alternative candidates and select the best option, the multi-criteria method requires four phases—(i) defining a set of attributes that most influence the choice of the alternative, (ii) a decision matrix, (iii) weighting of the criteria using an appropriate method, and (iv) evaluate alternatives by classifying them from best to worst.

It is possible to find several examples of the application of the Multi-Criteria Decision Aid (MCDA) methods [8] for the conservation and reuse of cultural heritage [10–15]. From an analysis of the main scientific literature with applications for MCDA in decision problems regarding cultural heritage, emerges the importance of the use of combined methodologies [16–21], which are particularly useful for dealing with complex situations [22–24], and are capable of structuring more inclusive decision-making processes aimed at a greater sharing of choices [25–27].

Many applications of MCDA are in the domain of cultural heritage. From the point of view of field application, the analysis shows that there are works in different research domains and that few works exist for the evaluation of cultural assets with reference to the strategies for their reuse [28]. As regards applications in the literature, multi-criteria analysis provides a theoretical and operational framework useful for supporting the design and implementation of public policy strategies in complex contexts with different values at stake and often conflicting [11–13]. The robustness of the multi-criteria methodologies is due to a series of main factors—heterogeneous criteria capable of managing both quantitative and qualitative information, multiple actors can be easily involved, the final aggregation of the different positions on adaptive reuse strategies [9–13], and the research for a solution of greater compromise that reflects the preferences of most points of view but also that of minority points [29].

This document addresses the issue of adaptive reuse of cultural heritage in a context of multi-criteria decision making and contributes to the cited literature. In detail, a decision aid system is developed, that identifies context-aware enhancement scenarios, oriented to the construction of new tangible and intangible values for the cultural heritage abandoned located in Southern Italy. What is proposed is an integrated evaluation model, based on the use of different multi-criteria methodologies, to define a classification between different scenarios and development strategies for unused historical heritage located in vulnerable contexts. The methodology is able to support the choice of an alternative reuse and to define a "shared strategy" based on a "bottom-up" approach.

The application to the case study concerns the potential reuse of six different historical fortifications (built at the end of the 18th century) located along the coasts of the Strait of Messina in Southern Italy, which are characterized by particular architectural and technical characteristics and by enormous spaces suitable for being redeveloped for new uses.

This requires the definition of the needs that must be met by the new functions assigned to the historical asset and the identification of the structural and cultural constraints to its transformation. From this perspective, a key problem is that of prior identification of the most suitable assessment methods for the decision problem.

The choice of methodology used here derives from the complementary relationship activated between the techniques of the Systems Thinking Approach [30–33] and the multi-criteria and multi-group evaluation methods, which allows us to outline a decision-making process that makes use of the specifics of each one in order to improve the knowledge of the context, to make explicit the preferences of the different stakeholders, to build shared visions of the future, and identify the preferable scenarios considering both the "qualitative" data and the "quantitative" data. The application of combined evaluation methodologies is justified by an increasingly widespread practice [30–37] aimed at overcoming the limits of each single method, to consider a multi-dimensional and plural perspective, to activate inclusive decision-making processes, and to improve the quality of the process oriented to the construction of shared choices.

The evaluation process implemented also integrates the contribution of different knowledge. It is based on the sharing of responsibilities between the different stakeholders and the concertation of choices through the complementarity between experiences and competences of different domains.

The main objective is to provide a decision support tool to design and implementation of adaptive reuse strategies of abandoned military heritage in a vulnerable context [37–39].

Furthermore, the combined use of multi-criteria techniques helps to classify unused adaptive cultural heritage reuse strategies and is able to support decision makers in implementing development strategies in vulnerable contexts. The application to the case study focuses on the potential reuse of some historical fortifications located along the coasts of the Strait of Messina. The results obtained show that the proposed model can be a useful decision support tool in contexts characterized by high complexity and that it is able to guarantee the transparency of the decision-making process for the construction of shared development strategies, in which it is necessary to highlight the elements that influence the dynamics of choice toward the best reuse project that reaches an optimal balance between conservation and usability [40–42].

In detail, the document is organized into five sections. Section 2 describes the methodological framework and the theoretical references of the techniques used. Section 3 presents the application to the case. Section 4 discusses the results obtained, while Section 5 traces the conclusions and identifies the research prospects.

### **2. Materials and Methods**

### *Methodological Background*

In the field of decision making, Multiple Criteria Decision Aid (MCDA) has become a widespread tool capable of dealing with problems involving more than one goal. A number of theorists [43,44] and empirical studies [45–47] have shown that the MCDA can manage the complexity of the entire process and through a participatory process can help avoid conflicts and make transparent decisions more effective, involving different points of view in the initial phase of the project/problem [48].

These decision support methods, in addition to the possibility of considering data of different nature (quantitative and qualitative), offer the opportunity to assign different weights to the criteria identified, manage conflicts between objectives, and deduce priorities among alternative options.

Many Multi-Criteria Decision-Making (MCDM) methods and techniques have been proposed by researchers in the past decades. Some of the most important ones are Weighted Sum Model (WSM) [49], Weighted Product Model (WPM) [50], Weighted Aggregated Sum Product Assessment (WASPAS) [51], Analytical Hierarchy Process (AHP) [52], ELimination Et Choix TRaduisant la REalité (ELECTRE) [53], Technique for Order of Preference By Similarity to Ideal Solution (TOPSIS) [54], Preference Ranking Organization Method for Enrichment of Evaluations (PROMETHEE) [55], Complex Proportional Assessment (COPRAS) [56], Visekriterijumska Optimizacija i kompromisno Resenje (VIKOR) [57], Multiobjective Optimization by Ratio Analysis plus Full Multiplicative Form (MULTIMOORA) [58], Additive Ratio Assessment (ARAS) [59], Evaluation Based on Distance from Average Solution (EDAS) [60], and Combinative Distance-Based Assessment (CODAS).

The choice of a particular multi-criteria method depends on the specific problem and user needs. It can be pointed out that the number of existing methods is already large enough and is still increasing, and choosing the "right" method is not an easy task. To facilitate the selection of the appropriate method for decision making, a list of quality criteria has been developed [9] that can be used to reveal the strengths and weaknesses of MCDA methods compared to their application to specific problems. For the evaluation of the individual methods, quality criteria were used to highlight differences in the characteristics of the methods presented. The study also highlighted that selecting the best method is not possible. In managing complex situations, all methods show similar performances with respect to the aspects and scales that can be considered.

All the MCDM methods mentioned above have advantages and disadvantages—their evaluation is not the purpose of this document. In this document, a hybrid method to manage multi-criteria decision problems has been proposed. Furthermore, in multi-criteria analysis models, determining the weights of the criteria is another of the key problems of the decision-making process as the weights of the criteria can significantly influence the outcome of the decision-making process. It is therefore clear that particular attention must be paid to the factors of objectivity in attributing the weights to the criteria.

A study on the advantages and disadvantages of individual criteria weighting methods is provided by Roberts and Goodwin [61]; it is also recognized that the values of the weights of the criteria are significantly conditioned by the methods used for their determination [61–65].

In the literature, it is possible to find a classification [53] of the methods for determining the weights of the criteria, but there is no agreement on the best method for determining the weights of the criteria. In the literature, most authors suggest a classification of the models for determining the weights of the criteria distinguishing subjective and objective models. Subjective approaches mirror the decision maker- s subjective opinion and intuition. With such an approach, the decision maker directly influences the outcome of the decision-making process, since the weights of the criteria are determined on the basis of information obtained from the decision maker or experts involved in the decision-making process. Objective approaches neglect the decision maker- s opinion and are based on determining the weights of the criteria based on the information contained in a decision matrix that applies certain mathematical models.

The best-known subjective methods for weighting the values of the criteria are the Analytic Hierarchy Process (AHP) method, the Step-Wise Weight Assessment Ratio Analysis (SWARA), the Best Worst Method (BWM), and the Full Consistency Method (FUCOM). Each of these methods has a wide application in the various areas of science, technology, and in solving real life problems.

It can be pointed out that, despite being a relatively old method, the AHP method is still used in a large number of publications. This confirms the conclusions of Zavadskas et al. [66] that the AHP method is the most commonly used method for determining the weights of the criteria and/or classifying alternatives.

The Analytic Hierarchy Process (AHP) method [67] is certainly one of the most commonly applied methods based on pairwise comparisons to determine the weights of alternative criteria and/or rankings [67]. The validation of the results in the AHP model is based on the degree of consistency, the value of which is limited to 0.10. In the AHP method, n (n−1)/2 pairwise must be performed. It should also be added that a large number of comparisons complicate the application of the model, especially in cases of a greater number of criteria. According to some authors [68], it is almost impossible to perform fully coherent pairwise comparisons in the AHP method if more than nine criteria exist. This problem is usually overcome by dividing the criteria into sub-criteria, which further complicates the model.

Therefore, this document, in order to classify different reuse alternatives [69–71], combines different multi-criteria methods in the different stages of the decision-making process in order to develop a "quick" tool to improve the quality and reliability of the decision-making process itself, especially in the planning decision phase. The Figure 1 presents the methodological framework of the decision support process used in the work.

**Figure 1.** Methodological framework of the decision support process.

In detail, to address the complexity of the decision problem under consideration and to support the choice about the highest and best adaptive reuse of defensive cultural heritage under study, an assessment has been implemented through the combination and integration of three multi-criteria methods: the MacBeth method [72–74], the Analytic Hierarchy Process (AHP) [68,75], and the Evaluation of Mixed data (EVAMIX) [75–79].

The combination and integration of multi-criteria methodologies [80–82] also allows the decision maker (especially in the planning phase) to classify the alternatives from the best to the worst, using a simple, effective and logical decision-making method.

In addition, the EVAMIX method was chosen as it is a fairly effective method compared to other approaches due to its simplicity of approach and the fact that it requires less time regarding the necessary mathematical calculations. To the knowledge of the authors of the article, this study represents the first experimentation relating to the combination of techniques of this type.

The application of the MacBeth evaluation method (Measuring Attractiveness by a Categorical Based Evaluation Technique) [72–74] was useful to evaluate the multidimensional impacts that the different functional choices (resulting from knowledge and participatory phases) can have on the general objective—increase the multidimensional productivity of the city. The elaboration of the data collected in the preliminary phase of knowledge and participation (focus groups and questionnaires) allowed to define the priorities regarding the functions to be included in the alternative scenarios. In the MacBeth method, the evaluation of priorities is based on pairwise comparisons and adopts an intervallic scale [8].

The Evaluation of Mixed data (EVAMIX) method [76–79], from a procedural point of view, allows the construction of an "m × n" evaluation matrix, characterized by "m" criteria and "n" alternative options, where the components are both qualitative rather than quantitative [83,84], whose lines express the performance of each alternative with respect to a specific criterion (see APPENDIX A Matrix of Evaluation). Definite 2.0 software was used to implement the EVAMIX method [85].

From a procedural point of view, the EVAMIX method consists of seven steps, for the analytical-mathematical procedure it's possible refer to the reference literature [65–68].

In practice, the EVAMIX method consists of the following passages [76–79]:


The first step is the construction of an evaluation matrix E, which is a mxn matrix, characterized by m evaluation criteria and n alternative options. The qualitative or quantitative data, contained in the matrix, express the performance of each alternative with respect to a given criterion.

The set of criteria is divided into two subsets—the set O is the set of ordinal criteria (qualitative) and the set C is the set of cardinal criteria (quantitative). This leads to two distinct evaluation matrices—EO (ordinal criteria/alternative options) and CE (cardinal criteria/alternative options).

Therefore, the differences between the alternatives can be expressed through two measures of dominance, the first based on ordinal criteria, the second on cardinal criteria. In particular, in order to construct the cardinal score of dominance, the components of EC are standardized by a common unit; this allows all quantitative criteria to be expressed on a scale from 0 to 1. The weights can be assigned to both criteria, both quantitative and qualitative.

In the second step, it is possible to calculate the "dominance scores" for all ordinal and cardinal criteria; these scores reflect the degree to which one option exceeds the other.

Subsequently (step 3), the dominance scores are standardized with the same unit of measurement to be comparable to each other, and two "standardized dominance measures" are deduced for all ordinal and cardinal criteria.

Thus (step 4), a "measure of overall dominance" is calculated for each pair of alternatives, which translates the degree to which an option dominates over another option.

Finally (step 5), starting from the overall dominance measure, the "evaluation scores" are calculated for each alternative: the result is a complete classification of the alternatives, where the optional alternative is the one with the evaluation score higher.

The AHP Method, developed by Thomas Lorie Saaty [68,75], structures decision making in the form of hierarchy. From a procedural point of view, this method consists of the following three phases:


In the first step, the AHP method helps to break down complex systems into simple structures. This simplification is made possible by a logical process that aims at the construction of adequate hierarchies.

The simplest hierarchy model consists of three levels. The first level is that of the main objective (called "Objective") of the decision problem; the second and third levels include criteria and alternatives. However, it is possible to develop more complex (i.e., multi-level) hierarchies including a number of sub-criteria. This means that the factors that influence the decision are organized in gradual steps from the general, in the upper level of the hierarchy, to the particular, in the lower levels.

In the AHP method, pairwise comparisons (i.e., comparison of elements in pairs with respect to a given criterion) are used to establish priorities (or weights) between elements of the same hierarchical level (step 2). These elements are compared in pairs with respect to the elements corresponding to the next higher level. This comparison translates into a matrix of pairwise comparisons.

To represent the relative importance of one element with respect to another, a suitable evaluation scale called the "Saaty scale" is introduced. It defines and explains the values 1 to 9 assigned to the judgments to compare pairs of elements of each level with respect to a criterion in the next higher level.

Pairwise comparisons are organized in appropriate matrices, since each of them is related to "priority vectors" (expressed on a 0–1 scale) and, when aggregated, they provide a complete classification of alternatives.

For pairwise comparisons, both quantitative and qualitative data can be used. As for the third phase, it should be noted that when comparing the elements, a certain degree of inconsistency can arise: in the AHP approach, a "coherence ratio" is calculated for each matrix of pairwise comparisons to verify the degree of inconsistency. A consistency ratio of 0.10 or less is considered acceptable; if this ratio is greater than 0.10, it is necessary to reformulate the judgments using new pairwise comparisons.
