**2. Approaches and procedures for data achievement**

communication within the social discourse, i.e., within a conversation [26]. In this line of significance, prices and conversations are subjected to the power/knowledge relation *a là* Foucault [27], and then to asymmetry affecting relations between media companies and users. In real markets, as well as in online conversations, asymmetric power and asymmetric information

Consequences for practices are many. Policy analysts and researchers must assist policymakers not in a technical way but in a political one, designing robust basic structure able to resist the constant shifting coming from political and social actors and lobbies, acting in the complexity of real life. Nevertheless, concepts used in formal models are useful criteria, beyond their mathematical rigor, for organizing the information found in case studies, and for evaluating policy alternatives and designing a specific governance in a specific

In this line, governance is worthy as an instrumental tool, because it provides the society (producers, consumers, and policymakers) with at least clear game regulations. In designing and managing practices of sustainable tourism, the ways the networks of governance do work at the local level are fundamental. Moreover, effective governance of a tourist destination can be self-improving eventually fostering participation and people commitment, and their perception of being immersed in a democratic selection of satisfying decisions. Providing a place with some tools capable of spreading information, discussion and learning, can make social negotiations effective. The functioning local governance is the central point for a serious approach to a credible sustainable tourism [29]. Thus, sustainable tourism is understandable as a procedure of regulations within the discourse of policy making and analysis, following

For all of this, meaning, misusing, and misunderstanding of words matter a lot in co-creating the enhancement of sustainable tourism, especially in the still "virtual" Adriatic-Ionian region

The objective of the paper is to illustrate the results of an explorative research of online conversations as retrieved from Twitter within a geographical area. Twitter microblogging is one of the social media used in the tourist sector, even its credibility should be contextualized [31]. The paper reports findings from a case study treating four tourist destinations located in the Adriatic-Ionian region, aiming at comparing the Eastern and Western coasts. The focus of the research was to analyze the meaning of *sustainable tourism* as possibly emerg-

Thus, the research questions are as follows: (i) how do suppliers, customers, and public agencies use Twitter for talking about sustainable tourism within the Adriatic-Ionian region? (ii) is there a difference in doing that between the two coasts of the Adriatic Sea? (iii) are public institutions using the web for designing and promoting their policy of sustain-

produce speculations and adverse selection behavior [28].

110 Mobilities, Tourism and Travel Behavior - Contexts and Boundaries

the constructivist postmodern vision [30].

**1.2. Objective, focus, and research questions**

ing from the texts produced by social media users.

and the specific area of social media.

able tourism?

place.

The approach of the research was explorative aiming at testing the possibility to use analysis of online conversation for delineating the meaning of *sustainable tourism* emerging from web users, including policymakers, within the Adriatic-Ionian region. For this purpose, the definition of the region was borrowed from European official documents [32]. In the region, four main tourist destinations of similar tourist appeal and dimension have been chosen. Singular destinations were selected based on the criteria of past reputation, established notoriety, the level of the main tourist indicators, such as tourism intensity per residents and area [33]. Thus, Cesenatico and San Benedetto del Tronto in Italy, and Makarska and Split in Croatia have been named as the most representative places, even considering all of them are destinations awarded by the Blue Flag program.

For each of them, the mass of themed texts retrievable from Twitter was collected during the last seasonal peak period, namely from July 15th to August 31st, 2016. Text analysis [34] of tweets can help to give the meaning of *sustainable tourism* as emerging from texting of actual actors. The definition of sustainable tourism given by a reputed eco-label tool was considered a useful proxy. For this, the Blue Flag program for beaches and marinas [35], run by the Foundation for Environmental Education (FEE) has been chosen and some sketches are given here following.

The Blue Flag program was started in France in 1985 and it has been operating in Europe since 1987, and outside Europe, since 2001, when South Africa joined. Today, Blue Flag has become a truly global program with an ever-increasing number of countries participating:

"The Blue Flag program promotes sustainable development in freshwater and marine areas. It challenges local authorities and beach operators to achieve high standards in the four categories of: water quality, environmental management, environmental education, and safety" [35].

For this, the choice is properly in line with the aim to consider *sustainable tourism* within the *marketing discourse*, as put in the previous paragraph. It appears valid to consider the Blue Flag program as a credible indicator of communicated sustainable tourism, and the above four categories as benchmarks for selecting keywords to be detected.

On the side of social media, the rationale for choosing Twitter is the following. This online medium is a huge deposit of pieces of information, being a website with 313 millions of users, 1 billion of accesses per month on websites embedding tweets, having 82% of mobile users, 3860 employees, and 35 offices worldwide, while having 79% of accounts registered abroad the US, 40 languages used, and 40% of technicians on total employees [36]. Even though at the present time, Twitter is reputed in a declining phase, it still has the capability of gathering content produced on other websites, such as Instagram, Facebook, and Tumblr and allows, even with some limitations, to survey user-generated data.

In order to optimize the process of gathering data, the web-based free service IFTTT (www. ifttt.com/wtf) was used. It allows connecting different online services furnished by other web tools, by using such a conditional recipe as "If This Then That"––the acronym of the service. Data have been downloaded in a spreadsheet stored on Google Drive adding a row whenever the designed "recipe" matched a new web-generated content. For instance, if someone did tweet the word "Cesenatico," the corresponding text was recorded, with date, hour, and user reference, in a dedicated spreadsheet. **Table 1** reports surveyed places and alternative location terms (tags) for selecting texts from Twitter. Alternative terms are related to common typing modes and languages of web users. **Table 2** reports different languages and keywords.

A geo-referenced condition was also given to IFTTT, namely if someone tweeted keywords within a 5 km radius area from the center place, the text was registered in a separated spreadsheet. Thus, data relating to a radius area from all the destinations have been collected.

Case sensitive and with and without hashtag (#) have not been used, accepting the burden of collecting fuzzy data and clearing them in the subsequent phase of analysis, aiming at not excluding any typing mode in texting.

Text analysis aimed at individuating the words displayed in **Table 2** when used in tweets and understanding the expressed sentiment. It seems important to remind that the searched terms are actually the Blue Flag criteria for defining sustainable tourism and the text mining was not robotized but manual.


\* The term #split was added in order to avoid confusion with the English "to split." However, a certain "white noise" remained, due to the movie of M. Night Shyamalan, *Split*, featured just on July 27th, which produced a lot of online conversations.

**Table 1.** Places and alternative location terms for selecting texts from Twitter.

Analysis of Online Conversations for Giving Sense to Sustainable Tourism in the Adriatic-Ionian Region http://dx.doi.org/10.5772/intechopen.70371 113

Data have been downloaded in a spreadsheet stored on Google Drive adding a row whenever the designed "recipe" matched a new web-generated content. For instance, if someone did tweet the word "Cesenatico," the corresponding text was recorded, with date, hour, and user reference, in a dedicated spreadsheet. **Table 1** reports surveyed places and alternative location terms (tags) for selecting texts from Twitter. Alternative terms are related to common typing modes and languages of web users. **Table 2** reports different languages and keywords.

A geo-referenced condition was also given to IFTTT, namely if someone tweeted keywords within a 5 km radius area from the center place, the text was registered in a separated spreadsheet. Thus, data relating to a radius area from all the destinations have been collected.

Case sensitive and with and without hashtag (#) have not been used, accepting the burden of collecting fuzzy data and clearing them in the subsequent phase of analysis, aiming at not

Text analysis aimed at individuating the words displayed in **Table 2** when used in tweets and understanding the expressed sentiment. It seems important to remind that the searched terms are actually the Blue Flag criteria for defining sustainable tourism and the text mining

> Rivieradellepalme Sambenedetto Sanbeach Sanbenedetto

Sanbenedettodeltronto

Macarsca Makarskariviera Makarskarivijera

Spalato Splitriva Splitriviera Spljet #Split\*

The term #split was added in order to avoid confusion with the English "to split." However, a certain "white noise" remained, due to the movie of M. Night Shyamalan, *Split*, featured just on July 27th, which produced a lot of online

**Place Alternative location terms (tags)**

San Benedetto del Tronto San benedetto del tronto

Cesenatico Cesenatico

Makarska (Macarsca) Makarska

Split (Spalato) Split

**Table 1.** Places and alternative location terms for selecting texts from Twitter.

\*

conversations.

excluding any typing mode in texting.

112 Mobilities, Tourism and Travel Behavior - Contexts and Boundaries

was not robotized but manual.



**Table 2.** Languages and words used (keywords) for text analysis.
