Online Place Branding for Natural Heritage: Institutional Strategies and Users’ Perceptions of Mount Etna (Italy)

: In recent years, tourist destinations and strategies of place branding have been facing new challenges owing to the di ﬀ usion of Information and Communications technologies. Smart devices can give tourists / prosumers the possibility to co-create and share their travel experience to the point to inﬂuence the destination web reputation and, consequently, its digital place image and branding. Furthermore, new technologies can be also used as e ﬀ ective analytical tools to scrutinize the role of online co-created narratives in inﬂuencing the web reputation of a speciﬁc tourist site, natural heritage included. The study focuses on the online destination image of Mount Etna, an active volcano located in Southern Italy inserted in the UNESCO World Heritage List since 2017. The web reputation of this natural heritage site has been analyzed through a twofold methodology: a manual online content analysis and a software-based Sentiment Analysis methodology. The paper highlights the crucial role of new technologies both as tools of analysis in tourism and heritage studies and as “catalysts” of e-narratives able to inﬂuence place images. In so doing, the research aims at providing other researchers and policy-makers with new theoretical and methodological insights about the challenges and potentialities of smart technologies in exploring the online place image, thus contributing to a novel conceptualization of place branding through the theoretical / operational framework of Place Branding 3.0. In particular, the mixed-method approach represents an innovative framework insofar as it provides an in-depth evaluation of users’ online perceptions both at the “micro” scale—at the level of contents, through the manual content analysis—as well as at the “macro” scale, thanks to the software-based Sentiment Analysis methodology.


Introduction
Over the last few years, place branding strategies have been widely used at the different scales in order to enhance the tourist attractiveness of heritage sites and destinations in a growingly competitive tourism sector.
As several scholars put it [1][2][3][4], while the mobilization of marketing techniques to "sell" places dates back at least to the 19th century, place branding is a rather new approach which mobilizes a wider repertoire of strategies and practices including coopting local stakeholders, creating internal consensus and attracting exogenous resources. The objectives of place branding overlap with the "traditional" aim of "selling" the places for tourist reasons through a multifaceted and complex image-building strategy.
However, image-building dynamics and place branding strategies have been completely upset by the advent of the Social Web and the diffusion of smart technologies. aimed at enhancing a wider repertoire of characteristics of places, perceived credibility and authenticity included [4,15]. As Govers highlights [14] "In a globalized world, more and more places compete more intensely, partly because, with increased tourism, migration and the global reach of media and technology, markets come into contact with places more frequently and at different levels, and therefore the 'corporate reputation' of places is becoming more important, particularly because of the product brands being inseparable-that is, places are automatically dealing with a brand extension strategy".
Since reputation derives from a complex tangle of strategies and symbolic actions, marketing is not the appropriate tool to support it. Furthermore, another element for an effective place branding is internal branding, which means targeting all the branding actions also to internal stakeholders, namely the public, private actors, civil society who have to "live the brand" [16] and foster a common sense and identity in order to make it effective. Although having traditionally received scarce attention in place branding literature, in recent years internal stakeholders have taken on a central role in creating collective place branding strategies [17]. The relationship among heritage enhancement, local communities, tourism and place branding has been strengthened in the last two decades, encompassing a variegated set of strategies finalized to select targeted "narratives" about a destination and consequently attract exogenous resources (capitals, visitors, inhabitants) in order to challenge increased levels of regional competitiveness [18,19].
Several studies about place branding "have mainly considered place branding as a "top-down", patronizing strategy, whereby marketing and management professionals, paid for by stake-holders, select ideas that then are translated into graphic icons, policy documents, internet portals, and tourists' and locals' opinions" [20] (p. 155). Although different forms of bottom-up participative or even performative place branding have recently emerged [21], the process of brand construction always entails a selection of narratives which cannot elude any longer the growingly pervasiveness of new technologies. This implies that a place branding strategy should be supported by offline and online actions in order to build a homogeneous destination image ranging from the choice of a coherent logo to a multichannel strategy.
Revisiting Urry and Larsen's seminal work [22], we propose the conceptual/operational framework of Place Branding 3.0. Following the authors, who significantly renamed their book renamed The Tourist Gaze 3.0, we agree with the need to "rethink the concept of the tourist gaze as performative, embodied practices, highlighting how each gaze depends upon practices and material relations as upon discourses and signs" [22] (pp. [14][15]. Whereas, as Dinhopl and Gretzel put it [23] (p. 129), "the mediation of the tourist experience by varying technologies has always been part of the tourist gaze literature", the advent of Web 2.0 has recently shaped a new "mediated" gaze, insofar as new mobile communication technologies enable tourists to be connected to a larger online social network. Tourists "take on a dual position, where their vantage point is once from their point-of-view, and once from the point-of-view of their social media audience. For the tourist gaze, this dual position means that tourists are actors and narrators at the same time" [23] (p. 129). So, they are often induced to "narrate" their travel experiences by sharing online contents (both textual and visual) while they are living them.
Whereas co-creation of the tourist experience has always been regarded as a crucial element to construct narratives about places [24], nowadays the paradigm shift-implying simultaneous multiple media channels and interconnected communication flows-has been accelerated by the ubiquitous connectivity provided by smartphones [25,26].
Being contemporary tourism "emblematic of a new mobility nexus", not only tourists usually remain "electronically" connected with their routine world even during trips owing to mobile devices, but also their tourist experiences are increasingly mediated by new technologies [27][28][29][30][31]. In effect, "storytelling is principally a post-consumption activity. Consumers present stories to others to relay memories of events and activities as well as their significance in terms of identity ( . . . ). With the continued development and uptake of social media platforms, tourism stories are increasingly presented in public domains through travel blogs. Blogs have been described as unleashed consumers' narratives". [32] (p. 2) As a result, evaluating "naturally occurring narratives" provides new insights about the ways digital travelers transform traveling in an evocative experience and, consequently, concur in re-territorializing the meanings of a place.
As Albanese [33] puts it, digital storytelling-namely a web-based storytelling-is everywhere, and it has the capacity of telling someone a story through an emotional approach incorporated in a prototype frame, where universal elements can be easily identified by everyone. This "emotional" approach is referred both to co-created narratives shared by online users and to those developed by institutional actors to build and effective place branding strategies, embedded in the wider framework of smart tourism.

The Role of Smart Tourism in Shaping Online Place Narratives
As above-mentioned, while place branding strategies have been completely transformed by the advent of the Web, new technologies have been transforming the tourism sector as a whole.
As Katsoni et al. [34] highlight, being the tourism sector a key driver in several national and regional economies, it is crucial to explore its latest developments, mainly triggered out by the unprecedented role played by smart technologies in tourism management, hospitality, and destination branding.
The growing convergence between smart technologies and tourist experiences have led many scholar to define "smart tourism".
It can be described as "tourism supported by integrated efforts at a destination to collect and aggregate/harness data derived from physical infrastructure, social connections, government/organizational sources and human bodies/minds in combination with the use of advanced technologies to transform that data into on-site experiences and business value-propositions with a clear focus on efficiency, sustainability and experience enrichment".
With this exhaustive definition, Gretzel et al. [35] (p. 181) provide new theoretical insights about smart tourism, insofar as the authors identify three distinctive levels of smartness as well as the differences with e-tourism (Table 1). According to the Authors, smart tourism is a further step in the process of growing digitalization of the tourist experience based on an extensive use of new smart technologies not only in the pre and post travel phase, but also during the on-site experience. Furthermore, the interactivity of e-tourism, strictly dependent on the Web 1.0 phase, has been replaced by the technology-mediated co-creation due to the growing pervasiveness of social media and mobile devices.
Smartness encompasses three different levels: • Smart tourist destination, regarded as specific dimension of the smart city framework, based on state-of-the-art technologies and integrated infrastructures making the destination accessible, interactive, sustainable; • Smart experience, which entails a technology-mediated experience producing real-time monitoring, information aggregation, travel personalization, ubiquitous connectedness trough mobile devices, context-awareness and active co-creation of online contents for tourists/prosumers; • Smart business, which refers to a multilayered ecosystem that permit the creation and exchange of tourist resources as well as co-creation of tourist experience. According to Buhalis and Amaranggana [7] the business component of smart tourism encompasses a dynamic interconnection among stakeholders, involving public-private collaboration, in addition to the digitalization of core business processes and organizational agility [36].
As a result, an exhaustive definition of smart tourism should span three layers of the three above-mentioned components: the smart information layer finalized to collect data; the smart exchange layer aimed at promoting interconnectivity; a smart processing layer supporting the evaluation, visualization, integration of (big) data.
According to another definition of smart tourism, it encompasses the use of smart devices, mobile and Internet-connected interfaces spatially localized (such as GPS or Google Earth), that allow tourists to orient themselves [9]. Secondly, it implies that tourists become co-creative active producers who can foster civic engagement at the local level, increased levels of interaction with local population and/or other tourists, in addition to increasing awareness about the destination, in so doing stimulating local sustainability. Finally, smart tourism entails a multi-way, multichannel approach as a new interface at the convergence between the two dimensions-the physical and the virtual.
Thus, smart tourism incorporates the implications of a growingly mobile world: "interactive travel has a lot to tell us about this changing social world, and especially about the way social life has become wrapped up in technologies of moving and communicating" [9] (p. 7). Owing to the ubiquitous connectedness and the relatively low expenditures in terms of time and costs, Web 2.0 has lately shaped patterns and practices of tourist experiences. Social media have become "the most powerful force driving travel planning and decision making, playing a crucial role in travelers' overall travel experience. Travelers now have access to almost unlimited travel information created by other travelers. In fact, travel content created online is considered to be more credible and trustworthy than reviews from professionals or marketer information" [37] (p. 2).

Mount Etna in Italy: Our Case Study
For this paper, we adopted a case study approach [38] which provides some predictions to be texted in other contexts of study.
Mount Etna is one of Italy's main tourist iconic attractions, an outstanding example of ongoing geological processes and volcanic landforms as well as one of the most active volcanoes on Earth, the highest active in Europe at approximately 3350 m. It is in a state of continuous eruption for half a million years. Its lower slopes have been shaped by humans for years to take advantage of its fertile soils devoted to extensive agriculture, terraced vineyards, groves of chestnut, apple and hazelnut trees. Lava has carved more than 200 caves, and lava-stone is used in many fields, ranging from art and craft to construction.
Apart from supporting relevant terrestrial ecosystems including endemic flora and fauna, Etna is a veritable natural laboratory for the study of ecological and biological processes [39].
In 2013, UNESCO recognized Mount Etna as a World Heritage Site, a huge 19237 ha area including the most strictly protected sections of the "Parco dell'Etna", the Etna Regional Natural Park established by a regional law in 1987 based on four different levels of protection (from the A-level integral protection to C and D anthropic zones), which includes 20 municipalities.
The complex multi-functionality of Etna Park highlights different land uses which can be conflictual not only within the agricultural sector (marked by an increasing clash between traditional cultures, such as citrus fruits and wine, and innovative cultures such as the tropical fruits) but particularly with reference to non-agricultural activities: "clashes of opinion are well known between naturalists, farmers, agro-industrial industry, tour operators, traders, urban planners, and so forth, relating to the allocation of human activity. The situation is even more complex when these problems surface within protected areas, where the protection of the natural environment and the use of resources must find a sustainable solution to coexist without stifling human activities that are subject to environmental restrictions". [40] (p. 1453)

The Online Analysis
The empirical research aimed at evaluating the online place image of the Mount Etna and the related place branding strategies through a twofold methodology which included two phases: the first one based on a manual content analysis, the second one focused on the Sentiment Analysis methodology. The in-depth description of methods is provided in the following paragraphs.

The Manual Content Analysis
The in-depth manual analysis, carried out between 2017 and 2020, encompassed the manual evaluation of both institutional strategies of online place branding and users' generated contents about Mount Etna as a tourist destination. The first stage was focused on an in-depth analysis of online place branding strategies based on a key-word search in search engines, in order to evaluate to what extent an "Etna Brand" could be detected both as a textual and a visual content (logo). Furthermore, the official online strategies were scrutinized in official websites and social media managed by the Etna regional Natural park.
We analyzed the institutional online place branding strategy revisiting the assessment model developed by Fernández-Cavia et al. [41] who developed a Web Quality Index (WQI) for official tourist destination websites, and we applied it to the overall online place branding strategy. So, we evaluated if the official multimedia channels aim at: persuading potential tourists; providing accurate and timely information for potential and/or current tourists ranging from accessibility to accommodation; conveying the basic place values of the destination in terms of brand equity; selling services and products; providing a platform to share experiences among destination managers and tourists and particularly among tourists themselves through the direct links to social media channels.
In a second stage, a sample of comments and reviews published on Tripadvisor was selected with the aim of evaluating user's Web narratives by comparing them with narratives retrieved from the most relevant Facebook pages and/or communities as well as travel blogs, identified through a "listening" research phase according to the number of members/followers, likes, KPI and so forth.

The Sentiment Analysis Methodology
As above-mentioned, new technologies have caused meaningful changes in all scientific fields, especially in Geography. Ubiquity of media, notably new media, provides top down and bottom up generated contents which can be produced everywhere and at any time. An increasing use of new media fosters an overproduction of texts and images that puts a greater emphasis on territory. From the perspective of Media Geography, territory is understood both as the place in which texts and images are produced and equally as the place narrated by texts and images. Larger volumes of accessible web generated contents often have an extremely strong influence both in and over territorialization processes. This epistemological change is highly significant and calls into question very crucial issues such as the new dynamics of social relationships, growingly embedded into the virtual dimensions, and the role of subjectivities in reshaping place perceptions.
In the current over-communicative context, territory (tourist destinations included) becomes a subject increasingly narrated online to the point to be re-configurated by new meanings and semiotic suggestions.
As a result, digital communication has a strong impact both in territorial storytelling and representation, by posing new theoretical and practical questions. In recent years research has focused on this theoretical-methodological framework to represent and deconstruct territorial complexity.
Our question is: to what extent are these new technologies of representation and narration influencing the creation of a sense of a territory and consequently its place branding?
In tourism studies, the huge use of digital technologies and social networks is generally aimed at sharing and exchanging information in digital places such as Facebook, Instagram, Twitter, and mostly through booking platforms, review sites and online blogs. The information gathered in these digital places help sharing travel information and descriptions of places but they can also convey emotions, experiences, sense of place and feelings. This web-based information is powerful because it can influence people's travelling choices insofar as users often re-build on their mind the sense of a place after having read online contents.
So, we decided to analyze the "sentiment" of Mount Etna, that is to say the online opinion emerging from web-based co-created contents, with a specific technique, the Sentiment Analysis methodology (SA) [42].
The semantic software specifically used for this study is App2check which includes several steps of analysis. It usually includes a first stage of manual "listening" of online "conversations", namely a manual content-based opinion mining which evaluates the online exchange of opinions among users.
This is aimed at understanding the general "sentiment" (opinion) on a given topic. By taping a set of key-words related to the research topic in a Web or social media search engine, the manual "listening" can explore the online conversation "ambiances", so that the researcher can build a sort of "online conversations framework". For this paper, we selected a corpus of 13,586 online textual and video sources (with a high number of comments for each social media source) retrieved from the first ten pages of Google. We selected Google as a search engine because of its influence: it is the most popular search engine and so it is easier to find chatroom of social media such as YouTube, Instagram, Facebook, Twitter and other forum rooms selected by topics. Using this method, we also "listened" (monitored) the most popular pages on the Web insofar as they represent the "conversations" taking place into the Web. These conversations, and the social networks in which they take place, are selected through the use of the following keywords: "Etna Turismo", "Etna Tourism", "visit Etna". After the tenth online page there are useless links insofar as they do not concur in creating the so-called e-WOM, electronic word of mouth. Thus, in Google we selected the conversation "ambiances" until the tenth page. Eventually, the social media from which we retrieved the data were Facebook, YouTube and Instagram, because they are those containing the greatest number of interactions among users.
After having retrieved the comments from social media selected for the analysis, we were able to go on with the next step of the semantic analysis: a process of splitting the sentences operated by the automatic software by separating the sentence into many pieces, called snippets, based on the number of adjectives or feelings expressed. Subsequently, the topic extraction took place. The most commonly discussed topics were isolated from the overall corpus of data. Topic extraction provided the so-called tags clouds, which are able to synthetically represent the polarization of online reviews and social media texts selected in the first phase of the research.
After having collected and selected the snippets and single words to be classified, the evaluation of sentiment took place. The sentiment range is usually translated into a numeric scale between 0 and 5, where zero indicates the most negative sentiment and 5, on the contrary, indicates the most positive sentiment.
Finally, the software provided some dashboards that connect the sentiment on different topics to a geographic location, the volume of comments and the content of online opinions. From the dashboard onwards, it was up to the researcher to extrapolate the conclusions of the semantic data retrieved by integrating them with a manual content analysis.
The sentiment analysis algorithm used is based on Breen's work [43]. Sentiment scores were assigned to sentences by calculating the difference between the number of words containing a positive sentiment and the number of words having a negative meaning.
In general, the positivity of a sentence is assumed when the score value is higher than zero, the negativity is assumed when the score value is below zero and neutrality when the score is equal to zero. Owing to the limitations of the software in assuming unambiguous negative or positive scores for sentiment values around zero, we decided to assume a positive sentiment when the score is equal or higher than 2, whilst negative sentiment with the score equal or lower than −2.

The Manual Content Analysis
As far as the manual content analysis is concerned, Etna reveals such an iconic territorial identity to be easily identifiable as a tourist site, in spite of the still existing gaps in terms of institutional place branding strategies. The key-word research in texts and images search engines has stressed the fragmentation of "Etna" as a brand, both as a stylized logo and as a name. More than 120 logos using "Etna" as a textual and/or visual components have been detected, ranging from private food and wine or tourist companies to cultural associations or institutions. As a heritage site, Etna has two official logos: the official logo of the Regional Park, which is also the managing body for the UNESCO site, and a newly-created logo as a UNESCO site (Figure 1). Finally, the software provided some dashboards that connect the sentiment on different topics to a geographic location, the volume of comments and the content of online opinions. From the dashboard onwards, it was up to the researcher to extrapolate the conclusions of the semantic data retrieved by integrating them with a manual content analysis.
The sentiment analysis algorithm used is based on Breen's work [43]. Sentiment scores were assigned to sentences by calculating the difference between the number of words containing a positive sentiment and the number of words having a negative meaning.
In general, the positivity of a sentence is assumed when the score value is higher than zero, the negativity is assumed when the score value is below zero and neutrality when the score is equal to zero. Owing to the limitations of the software in assuming unambiguous negative or positive scores for sentiment values around zero, we decided to assume a positive sentiment when the score is equal or higher than 2, whilst negative sentiment with the score equal or lower than −2.

The Manual Content Analysis
As far as the manual content analysis is concerned, Etna reveals such an iconic territorial identity to be easily identifiable as a tourist site, in spite of the still existing gaps in terms of institutional place branding strategies. The key-word research in texts and images search engines has stressed the fragmentation of "Etna" as a brand, both as a stylized logo and as a name. More than 120 logos using "Etna" as a textual and/or visual components have been detected, ranging from private food and wine or tourist companies to cultural associations or institutions. As a heritage site, Etna has two official logos: the official logo of the Regional Park, which is also the managing body for the UNESCO site, and a newly-created logo as a UNESCO site ( Figure 1) By taping different multilingual key word associations (such as Etna tourism, visit Etna etc.), any tourist portal has been found, except the website of the Regional park.
As the Table 2 shows, the official online place image is controversial. As far as the official website is concerned, first it is halfway between an administrative portal and a tourist one, including both "internal" documents (calls for tenders, committee reports and so forth) and tourist-targeted information (Figure 2). On December 2020, several sub-sections are under construction and even the news section is not updated. The section containing tourism information ("Environment", "Routes" etc.) has an out-of-dated lay-out, characterized by an over-presence of textual contents (Figure 3). Any booking section or multilingual sub-sections has been detected. By taping different multilingual key word associations (such as Etna tourism, visit Etna etc.), any tourist portal has been found, except the website of the Regional park.
As the Table 2 shows, the official online place image is controversial. As far as the official website is concerned, first it is halfway between an administrative portal and a tourist one, including both "internal" documents (calls for tenders, committee reports and so forth) and tourist-targeted information (Figure 2). On December 2020, several sub-sections are under construction and even the news section is not updated. The section containing tourism information ("Environment", "Routes" etc.) has an out-of-dated lay-out, characterized by an over-presence of textual contents (Figure 3). Any booking section or multilingual sub-sections has been detected.    With regard to the social media channels, the official website includes three links to a Facebook Page, a Twitter account and a YouTube Channel. Apart from having a scarce amount of contents, the two last platforms are not updated since years (Twitter since 2015 and YouTube since 2016), while the link to the Facebook page is not active.
However, by taping "parco dell'Etna" on the Facebook internal search engine, an official page has been detected which results to be active since 9th November 2020 ( Figure 4): this means that the old official Facebook page has been replaced by a new one, without updating the links into the website.  With regard to the social media channels, the official website includes three links to a Facebook Page, a Twitter account and a YouTube Channel. Apart from having a scarce amount of contents, the two last platforms are not updated since years (Twitter since 2015 and YouTube since 2016), while the link to the Facebook page is not active.
However, by taping "parco dell'Etna" on the Facebook internal search engine, an official page has been detected which results to be active since 9th November 2020 ( Figure 4): this means that the old official Facebook page has been replaced by a new one, without updating the links into the website.   With regard to the social media channels, the official website includes three links to a Facebook Page, a Twitter account and a YouTube Channel. Apart from having a scarce amount of contents, the two last platforms are not updated since years (Twitter since 2015 and YouTube since 2016), while the link to the Facebook page is not active.
However, by taping "parco dell'Etna" on the Facebook internal search engine, an official page has been detected which results to be active since 9th November 2020 ( Figure 4): this means that the old official Facebook page has been replaced by a new one, without updating the links into the website.  Although having been recently created, this new Facebook page is followed by 5468 followers (with 5360 likes), it is constantly updated (on average every two days) and based on interactive strategies of users' engagement, such as photo contests or calls to action. The Facebook page is connected with a newly-created Instagram account ( Figure 5): even in this case, in spite of the novelty, the account is based on the "pillars" of social media communication (engagement, emotional involvement and so forth). Furthermore, two apps have been recently created: one designed to provide up-to-date geo-based information and the other including videos based on sign language, thus promoting an inclusive travel experience.
Heritage 2020, 3, Firstpage-Lastpage; doi: FOR PEER REVIEW www.mdpi.com/journal/heritage (with 5360 likes), it is constantly updated (on average every two days) and based on interactive strategies of users' engagement, such as photo contests or calls to action. The Facebook page is connected with a newly-created Instagram account ( Figure 5): even in this case, in spite of the novelty, the account is based on the "pillars" of social media communication (engagement, emotional involvement and so forth). Furthermore, two apps have been recently created: one designed to provide up-to-date geo-based information and the other including videos based on sign language, thus promoting an inclusive travel experience. As far as the online users' perception is concerned, Etna has been ranked as the first tourist attraction in Italy according to Tripadvisor reviews in September, 2016 and the first over the 192 attractions of Sicily (October 2020). Mount Etna has collected 11,910 reviews (mainly in Italian-4169-followed by English, French, Spanish, Russian), 8526 of which consider it as "excellent", 2492 "very good", 614 "medium", 178 "second rate", 100 "awful" (last access: October 2020). Obviously, reviews have been less numerous in 2020, probably due to the COVID-19 pandemics and the subsequent decrease in tourist flows, particularly from abroad.
In general, comments stress the moon-like atmosphere of the site, the breathtaking landscape from the "roof of the world", the unforgettable experience of walking on lava stone and the noises coming from the "bowels" of the earth. However, in spite of underlining the sublime spectacle of nature, "awful" comments are mainly focused on the high costs and disorganization of second rate tourist services (funicular railways to reach the highest craters, parking, guided jeep tours) and outof-date facilities and infrastructures: "Nothing to tell about the breathtaking beauty of the landscape, but prices are really out of control! Pay attention". (a visitor from Milan, Italy, September 2019). As far as the online users' perception is concerned, Etna has been ranked as the first tourist attraction in Italy according to Tripadvisor reviews in September, 2016 and the first over the 192 attractions of Sicily (October 2020). Mount Etna has collected 11,910 reviews (mainly in Italian-4169-followed by English, French, Spanish, Russian), 8526 of which consider it as "excellent", 2492 "very good", 614 "medium", 178 "second rate", 100 "awful" (last access: October 2020). Obviously, reviews have been less numerous in 2020, probably due to the COVID-19 pandemics and the subsequent decrease in tourist flows, particularly from abroad.
In general, comments stress the moon-like atmosphere of the site, the breathtaking landscape from the "roof of the world", the unforgettable experience of walking on lava stone and the noises coming from the "bowels" of the earth. However, in spite of underlining the sublime spectacle of nature, "awful" comments are mainly focused on the high costs and disorganization of second rate tourist services (funicular railways to reach the highest craters, parking, guided jeep tours) and out-of-date facilities and infrastructures: "Nothing to tell about the breathtaking beauty of the landscape, but prices are really out of control! Pay attention". The experience, as a visitor from UK stressed in September 2016, was defined "shockingly expensive" to the point to recommend other users do not waste their money: "You can pay 30€ to arrive to the first station or pay 70€ + 9€ to arrive to the second station and a tourist guide that you are forced to pay. What they don't tell you is that is enough to pay only the 30€ for the first station and then we can go walking until the Etna. This means that the 49€ euros extra they made you pay are not necessary. Also, this tourist guide you are force to pay is useless because he didn't say or explain anything".
(a visitor from Spain, January 2017) Apart from the expensive costs, many visitors stress the complete lack of professionalism of people working in restaurants and tourist facilities: "We were very much looking forward to our day going up Mount Etna. Boy, what a disappointment ( . . . )".
"What was most disappointing was no one had told us anything about the limitations on the tour. Quite the contrary, we were ready for a good hike up from where the buses would drop us. Instead, we got a 10-min walk, with virtually no elevation, to a view towards the coast line. That was it. We could've done the whole thing in 20 min. It was completely ridiculous".
"Another example of how this was a completely disappointing experience was the way it started. We had traveled for a good distance and at the base of the gondola asked for the bathroom. The one there was out of order. We were told to go down and use the one below at the restaurant. Of course, there, there was a long line. After the long wait, you got the opportunity of using a stall had no seat and no paper. You have to be kidding. This simply is not the way to operate any serious attraction".
"So, at the end of the day, we were over charged for a completely uninspiring visit and the facilities were gross. Guess what, we don't recommend it".
"At the very least, if you are to do Mount Etna, you must call ahead and see how much is open. I promise they won't tell you unless you ask".

(a visitor from Boston, May 2017)
Finally, the analysis was enlarged to Facebook non-official/informal groups and the blogosphere. The "listening" phase has led to the selection of Facebook communities and/or open pages with more than 1000 likes/members, analyzed between 2017 and 2020, after having discarded those related to commercial private companies or public institutions ( Table 3). The in-depth content analysis has underlined the use of different online contents, with a strong preference for videos and photos, while textual contents are mainly based on the celebration of the breathtaking landscape without any reference to the controversial issues highlighted by Tripadvisor users. The same trend can be detected in all the selected travel blogs (written in English or in Italian), where the awe-inspiring scenes of nature, vineyards and small villages included, are the main topics. In four blogs, Mount Etna is depicted as an off-the-beaten track destination, in spite of the thousands of tourists and the globally renown reputation, probably due to the harsh wilderness of an uncontested overwhelming nature. Even in this case, any negative aspect is underlined.

The Software-Based Analysis
The manual analysis of the Mount Etna online place image was integrated with a software-based methodology carried out through the App2Check software for Sentiment Analysis methodology, which provides useful information about online users' preferences. The software was set up to decode the contents in two languages that were regarded as the most meaningful sources of information, namely Italian for the analysis of online perception of domestic demand, and English to obtain an international overview of the results. The overseas component of demand is shown in specific dossiers [44] to be on the rise nationally, while in Sicily overseas demand is low but it is growing rapidly mainly from within the European Union.
After the "setting" above descripted, the software managed and organized the analysis' results into tag clouds and graphics, such as bar graphs.
The in-depth analysis of online opinions written in both languages has provided several useful observations. Image 6 represents the general tag cloud where all comments are included. Some recurring topics such as "Etna", "Etnapeople", "volcano", "excursions", "beautiful", "amazing", etc., can be easily identified from the tag cloud. Words are polarized on the same topic, so the tag cloud does not represent a lot of words: the most recurring words are represented in the tag cloud according to the times they occur in comments.
The sentiment is positive, as it is shown by the total value at the right bottom of the Figure 6, equal to 3.84. On a sentiment scale ranged from 0 to 5, the value 3.84 is above the average and therefore it can be considered more than positive. The tag cloud is based on another tool to evaluate the "quality" of the sentiment, namely the color: green-colored words are positive, red-colored words are negative and yellow-colored words are neutral. The more intense is the color of the words, the closer the sentiment is to the pole (positive or negative).
Heritage 2020, 3 FOR PEER REVIEW 14 Figure 6. Overall Tag Cloud for Etna during 2020. Source: our elaboration from App2check. Figure 6. Overall Tag Cloud for Etna during 2020. Source: our elaboration from App2check.
By scrutinizing the words from the color standpoint, it is clear that the red ones (which means related to adjectives with a negative sentiment) are completely lacking.
Furthermore, it is also very useful to compare these data through a diachronic analysis (Figure 7).  The chart of Figure 7 helps us to better interpret data. Negative comments refer to the first phase of the summer season, when the whole tourism sector was still suffering from the negative aftermath of the post-lockdown. Subsequently, the curve goes back to high rates throughout the month of August.
Inevitably, this leads us to ask ourselves a crucial question: if these are the data we obtained in an extremely exceptional phase on a global level, that is during the COVID-19 pandemic, what are the data related to Etna's natural heritage in the year prior to the 2020? Figure 8 shows us the trend of sentiment during the 2019 summer season. The chart of Figure 7 helps us to better interpret data. Negative comments refer to the first phase of the summer season, when the whole tourism sector was still suffering from the negative aftermath of the post-lockdown. Subsequently, the curve goes back to high rates throughout the month of August.
Inevitably, this leads us to ask ourselves a crucial question: if these are the data we obtained in an extremely exceptional phase on a global level, that is during the COVID-19 pandemic, what are the data related to Etna's natural heritage in the year prior to the 2020? Figure 8 shows us the trend of sentiment during the 2019 summer season. As you can see in Figure 8, which shows the chart related to the pre-pandemic season, the curve trend is more homogeneous insofar as there are no changes in the sentiment at all and the curve never drops below the green area that represents positive comments.
So, we scrutinized what were the most discussed topics in the tourist season prior to the pandemic. The Tag Cloud related to the 2019 tourist season, shown in Figure 9, highlights the centrality of the word Etna combined with positive feelings. As you can see in Figure 8, which shows the chart related to the pre-pandemic season, the curve trend is more homogeneous insofar as there are no changes in the sentiment at all and the curve never drops below the green area that represents positive comments.
So, we scrutinized what were the most discussed topics in the tourist season prior to the pandemic. The Tag Cloud related to the 2019 tourist season, shown in Figure 9, highlights the centrality of the word Etna combined with positive feelings. As you can see in Figure 8, which shows the chart related to the pre-pandemic season, the curve trend is more homogeneous insofar as there are no changes in the sentiment at all and the curve never drops below the green area that represents positive comments.
So, we scrutinized what were the most discussed topics in the tourist season prior to the pandemic. The Tag Cloud related to the 2019 tourist season, shown in Figure 9, highlights the centrality of the word Etna combined with positive feelings.  Sentiment is close to 4, with a value of 3.93, and it is therefore more positive than that recorded during the pandemic. This confirms our doubt about the first chart in Figure 7, since in that case we recorded some negative and neutral sentiment due to the pandemic and not to the destination itself.

The Online Content Analysis
As the case study has demonstrated, narratives and perceptions about a tourist destination can be shaped by an extremely variegated set of dynamics and travel experiences, apart from being deeply influenced by the characteristics of the online channel chosen to create and share contents.
As far as institutional place branding is concerned, Etna cannot be regarded as a truly smart tourist destination, since online institutional strategies have demonstrated to be too much fragmented to be categorized as smart as in Gretzel et al.'s perspective [35]. While other smart elements are missing (e-booking, e-ticketing, smart infrastructures and so forth), it is principally the online image destination that has resulted to be not coherently inserted in a comprehensive strategy.
By incorporating the empirical analysis into the place branding framework [14,16], it is evident that the institutional place-building construction does not pursue all the objectives that are regarded as essential [41], such as persuading potential tourists, providing them with a timely information and selling services and products. The manual analysis has identified some recent efforts to convey the basic place values of the destination in terms of brand equity through the emotional engagement of online users in social media such as Facebook and Instagram, and through the newly-created apps for smartphones. However, these new "channels" are too recent to be fully evaluated in this work, and they should be better connected in a more comprehensive strategy by including the official website as well as more innovative e-services.
While Mount Etna suffers from a fragmentation in terms of official place branding, it still remains one of the most celebrated attractions in Italy thanks to the "natural" online brand identity, as the content analysis of blogosphere and social media has demonstrated. In spite of also being ranked among the first attraction in Tripadvisor, negative comments underline some structural gaps depending of the strategies of tourist management, thus mirroring the increasingly crucial role of peer-to-peer reviews and online "world-of-mouth" (namely the creation of a web reputation thanks to the information shared and diffused into the Web) which reveal critical issues that are undervalued in the blogosphere or in social media communities.

Challenges and Potentialities of Software-Based SA Methodology
Online social spaces can be considered anthropologically dense. They are intersected with a wide range of relations of communication favored both by online social networks and networks occurring in real-word. In this scenario, the S. A. methodology is a relevant tool to examine how these relations support new destination branding strategies and/or reinforce an already existing destination brand. Perspective and current tourists use online communication to share opinions; these opinions influence territorial perceptions and imaginary and this new imagery can be often translated into new tourist choices. So, S.A. provides a new operational framework to predict (ex ante) and analyze (ex post) travel choices by inferring them through a software-based scrutiny of online sentiments and opinions.
As above-mentioned, e-tourism has become a very relevant niche in the tourist sector as a whole, and it makes necessary to develop coherent online destination branding policies for the promotion of tourist products and/or destinations also by exploiting the added value deriving from online users' comments and evaluations.
It is therefore crucial to use and adapt S.A. to explore how a territory is perceived. Regarded as a specific methodology within the opinion mining umbrella, SA is in its early stages, but it will be increasingly used to systematically gather and examine big data from the Net to provide hitherto unimaginable qualitative information [45]. S.A. applied to our case study has allowed us to understand how a natural heritage, excluded from specifically-tailored marketing policies, still has its own recognizability and it is narrated with positive opinions (sentiment).
Given the current transformations in the tourism sector, it was necessary to analyze our topic through a diachronic approach, by evaluating the sentiment about Etna before and after a fundamental watershed: the COVID-19 pandemic. The first phase of S.A., carried out with reference to the pandemic times, provides a perception about the tourism sector in Sicily; more specifically, in the Etna area less green (positive) words have been identified than in "not-pandemic" times because of the altered perception of travelling due to the COVID-led crisis. So, it was crucial to carry out a novel S.A. analysis referred to a period prior to the pandemic which has shown how, without unpredictable events (such as COVID-19), sentiment is devoid of neutral and negative opinions, and it is mostly based on positive values. Therefore, the software-based S.A. has highlighted how Etna topic is on average green, and consequently positive, in every stage analyzed.
The case-study demonstrates to what extent new media have raised the importance of common feeling, sharing and exchanging experiences, also when these communication-based relations shape a "sense of place" capable of influencing users' imagination and, subsequently, tourists' expectations. The more positive is the online narrative, the better is the territorial imagery that derives from it. As the case study has highlighted, the extremely positive narrative elements can arise an equally positive imaginary in users' minds, thus triggering a positive spiral about the sense of place thanks to e-wom, namely online word of mouth, which amplifies the effect of digital messages thanks to the trust that e-users place in each other.

Discussion
Although the Web is accepted by scholars as a valuable resource and source of information to advance research on a variety of fields, including Geography, using web-retrieved data and a web-based methodology of analysis is not without its challenges. As Martì et al. [46] highlight, in spite of the advance in terms of collection times, data amounts, accuracy of geolocation marks and sample representativeness, online user-generated contents can lack consistency in the provision of a valuable set of geocoded data for each sample. Second, a lasting digital divide at different scales and dimensions can preclude a "carpet" sampling, since an amount of data depends on the ownership of a smartphone and the access to an Internet connection. Thus, it can be biased by several factors ranging from socio-economic status to political contexts and age. Third, the lack of transferability and representativeness in the information retrieved from social media can represent another limitation, in addition to the huge amount of dynamically-generated data sets, consisting of billions of observations, which can pose a number of challenges for geographers.
As far as the manual content analysis is concerned, the main advantages derive from the researcher's crucial role in sampling and identifying the corpus of contents to be analyzed in addition to providing an in-depth scrutiny from the qualitative point of view. Nonetheless, some limitations can be highlighted, such as the lack of a huge corpus of data.
On the contrary, the S.A.-based methodology can provide an in-depth evaluation of the sentiment expressed in a big amount of data, thus mirroring a wider representation of the phenomenon. However, as Sánchez-Rada and Iglesias [47] highlight, S.A. poses several challenges to natural language processing, caused both by the very nature of social media platforms (such as limited length of online opinions, for instance) or the specific features of human interactions in these kinds of media (e.g., short attention span, over-immediacy, slacktivism, use of jargon or abbreviations and so forth).
With regard to the Etna online image, the "three gap place branding model" developed by Govers and Go [2] identifies the main recurring gaps of place branding strategies, namely the divergences between product and identity, insofar as not always identity uniqueness is reflected in products; between product and experience, since the product performance can be disappointing owing to off brand implementation; and between experience and image, because of the different brand perceptions deriving from different socio-cultural backgrounds. By revisiting their perspective, we can assert that a multitude of gaps in the Mount Etna place branding strategies can be detected, in spite of the "tightness" of its web reputation and the recently emerged efforts in restructuring the online destination image trough apps and social media. As a result, the analysis can provide useful suggestions for building a multifunctional "smart" travel experience, which means to create an integrated, personalized, multifunctional tourist destination where offline and online strategies, narratives and discourses-both top-down and bottom-up-should be strictly interconnected in order to reinforce the already existing positive place "sentiment".
What is more, it is crucial to further problematize place branding both from the conceptual and from the operational point of view, going beyond the ubiquitous vision that identifies it as the panacea to attract exogenous tourists, investors or new residents. The more local communities are involved in place-image building processes, the more the place branding strategy is effective, insofar as it mirrors bottom-up expectations and local symbolic projections about places [48][49][50]. This is embedded in the so-called "right to the brand" [4,49], increasingly shaped by narratives coming from local communities. Following Vanolo [4], we can adapt his exhaustive overview about city branding to a branding strategy referred to a broader area, such as that one gravitating around Mount Etna. The author highlights to what extent the diffusion of business concepts-whether it be marketing or branding-to discourses on places mirrors the advent of neoliberal times, embedded in Harvey's entrepreneurial vision [51]. This involves that the economic rational informs all the spheres of human and social life: "In the case of cities-and places in general-it means that towns and regions are more and more assumed to function, under many perspectives, as products or services that can be packaged, promoted, sold and consumed through the market: this is the classic hypothesis of the commodification of places". [4] (p. 28) So, in order to avoid the risk of commodifying places through mere market-oriented strategies, it is crucial not to consider tourists as passive consumers, insofar as they may develop attachment to a destination because of a complex repertoire of elements, ranging from basic needs to symbolic implications, owing to the growing pervasiveness of the tourist gaze [9,15,17,22,23]. As Vanolo puts it [4] (p. 50), "the pervasiveness of the tourist gaze makes us 'everyday tourists', meaning that the tourist gaze operates also in absence of mobility, as in the case of consumption of digital images through the Internet".

Conclusions
Following Bosangit et al. [32] (p. 2), who hold the view that "narratives are fundamental in the construction of tourism experiences", we stress the need to take into account the variegated repertoire of narratives emerging from territories in order to build a coherent online destination image. Particularly in place branding strategies, both top-down institutional narratives and bottom-up narratives shaped by tourists and/or wannabe-tourists should be considered as crucial. The difference with the pre-Web tourism is given by the unmatched possibilities of continuously re-writing and sharing these narratives in every stage of travel (pre-travel dreaming/planning/booking phase; on-site-experience; post-travel phase) [14,16]. Place branding should consequently be activated in every above-mentioned stage as a comprehensive strategy aimed at both attracting and re-shaping tourist gazes, increasingly regarded as active prosumers (producers and consumers) of place images. The mixed-method approach used in this work can represent an innovative framework to explore some elements related to place image building by providing an in-depth evaluation of users' online perceptions both at the "micro" scale-at the level of contents, through the manual content analysis-as well as at "macro" scale, thanks to the software-based S.A. methodology. So, although the above-mentioned limitations, our research has demonstrated to what extent social media can impact the quantity of information available about places to the point to be applicable to wide areas such as the Etna region, which includes a complex system of multilevel territorial governance (i.e., municipalities, regional park, UNESCO). What is more, the complexity of Etna region is further emphasized by the overlapping-and often conflictual-territorial dimensions related to urban settlements, agricultural systems and tourist facilities, each with specific systems of actors. Although not always transferrable to other spatial contexts, the analysis of techno-mediated narratives provides crucial details about territorial perceptions which could be effectively used to make destinations more attractive. Thus, in spite of the limitations, the effectiveness of new technologies and the Web as sources of data has been widely demonstrated, particularly in the analysis of tourist destinations.
Finally, the case study provides new insights from the theoretical point of view, insofar it has emphasized the growing pervasiveness of users' generated contents in shaping place images and, consequently, tourist web reputation, which partially outweighs "fragmented" institutional online place branding.
From this perspective, place branding should develop into a new conceptual/operational framework-the Place Branding 3.0-which should incorporate new technologies both as tools of analysis of place image and as instruments growingly embedded in travel experiences and bottom-up processes of place image co-creation.