Facebook and virtual nationhood: social media and the Arab Canadians community

This article focuses on the study of online communities and introduces an empirical study of social media production involving an online group called “Arab Canadians”. The study builds on Anderson’s concept of ‘imagined communities’ and argues that Facebook provides the platform for an online nation in which users, whether Canadians or prospective immigrants, interact and exchange ideas about a country whose imagined concept varies from one user to another. Facebook here is a virtual nation that offers the community members an imagined sense of identity and belonging which they aspire to get. The results of the study revealed that the majority of comments carry highly positive sentiments towards Canada and its people, yet there is evidence that some comments are moderated. The study concludes that the Facebook administrator functions as a centralized gatekeeper who filters online chatter and leads the discussion to a certain direction. Building on the theory of networked gatekeeping, the study argues that vertical and horizontal flows of communication shape the online debate that takes place in this virtual space. Through a close analysis of these practices, the article sheds light on the role of social media in shaping online identities constructed around virtual nationhood.


Introduction
This research empirically investigates an online community by exploring its social media expressions which are regarded as a horizontal type of communication. The study attempts to understand the dynamics between horizontal and vertical communication in social media spaces, by analyzing the influence of a social media moderator on the discussions on a Facebook page. This aspect of vertical communication is linked to the theory of networked gatekeeping, while Anderson's concept of 'imagined communities' is used to highlight the way Facebook functions as a platform for virtual nationhood. These aspects are investigated through an empirical study of social media production involving an online group called ''Arab Canadians''. Before discussing the details, it is important to provide an overview of the Arab Canadian community living in Canada.
The Arab community living in Canada is regarded as a visible minority group and is known to have a 23.8% unemployment level which is the highest in Canada as of 2006 (Jedwab 2008). While the figures change over time, Canadians from Arab descent were estimated to be 348,000 according to the 2001 census (Census of Canada 2001) and this population continues to increase. For instance, Canada welcomed 34,567 Arabs in 2010 (Canadian Arab Institute 2013) and by the beginning of 2016 over 25,000 Syrian refugees were admitted into the country following the new Liberal government promise (CBC 2016). In Montreal and Ottawa, Arabs come second after Blacks as the largest ethnic group (19.7%), and they constitute the second youngest visible minority group, as their median age is 30.2. In addition, Arabic is considered to be among the top ten foreign languages spoken in Canada by ethnic groups (Statistics Canada 2011). According to the last Canadian census held in 2011, there were more than 750,000 Arab Canadians living in Canada mostly residing in Quebec and Ontario (Canadian Arab Institute Bulletin 2013). However, more details on Arab Canadians are not publicly available because of the scarcity of studies on some ethnic minorities and the fact that the mandatory long-form census was cancelled in 2011 and replaced by a voluntary short form version. Most importantly, empirical studies on Arab Canadian's social media use are rare; hence, this paper fills a gap in the literature by providing insight into the way an online community is formed and how its members interact with each other. In general, Arab Canadians have often been negatively viewed in public debates and media coverage, as they are mostly associated with terrorism, violence, and suspicion (Kashmeri 1991;Khalema and Wannas-Jones 2003;Hennebry and Momani 2013). Some politicians accuse Arab Canadians of not investing enough effort to integrate in the wider Canadian culture, while others believe that they suffer from racial profiling and forced marginalization in the multicultural society of Canada (Biles and Ibrahim 2002;Abu-Laban 2002).
Instead of examining what others think of Arab Canadians, the study reported in this article is focused on the way Arab Canadians view Canada and conceive their identities. The study draws on data collected from online interactions on a Facebook page that belongs to an Arab Canadian community. The article will present an analysis of the online interactions that occur among the members of the online community, as an illustration of horizontal communication flows, and interpret these in relation to the influence of vertical communication, namely the influence of the Facebook page administrator. The study found that the administrator acts as a moderator of debates to filter the online interactions, however, the networked audiences are also active in what is known as non-centralized flows of information. The article will expand on these topics, and provide an analysis of the role of social media in shaping online identities constructed around virtual nationhood.

Arab Canadians in the literature
To further understand the Arab Canadian community, the researcher examined the term ''Arab Canadians'' using the Google Books search tool (Thelwall 2009) to trace its first occurrences. It seems that one of the first references to Arab Canadians was published in the edited book by Hagopian and Paden that is entitled The Arab Americans: Studies in assimilation (1969) in which they are mentioned in relation to the Arab-Israeli conflict. Yet, based on another Google search conducted without the help of the same online tool, the researcher found another publication entitled Documents on Canadian External Relations: 1947 (Canada Department of External Affairs 1967) which made a reference to Arab Canadians in relation to Palestine. Using Google Ngram which examines book publications (https://books.google. com/ngrams), it becomes apparent that there was a sudden climax in the study of Arab Canadians around the year 1980 which could be related to the publication of Baha Abu-Laban's canonical work An Olive Branch on the Family Tree (1980) which offered a historical and cultural account of Arabs in northern America ( Fig. 1). At that time, Abu-Laban asserted that ''Arab communities have been studied less than most ethnic groups in Canada' ' (1980, ix). It seems that Abu-Laban popularized the term and several other works followed by citing it and expanding the research on Arab Canadians (See Fig. 1).
In comparison to another search term such as ''Muslim Canadians'', we find that the term ''Arab Canadians'' is more popular, despite the fact that the number of Muslim Canadians is much higher than that of Arab Canadians alone (See Fig. 2). Yet, when we compare the term ''Arab Americans'' with ''Arab Canadians'', we find a great disparity, as the number of publications on Arab Americans is much higher than that on Arab Canadians, especially since the study of Arab Americans started much earlier than that on Arab Canadians (See Fig. 1). If one examines specific Arab Canadian groups, (s)he finds much fewer works.
Based on a thorough literature review, it seems that previous research on Arab Canadians has not focused on the use of social media, despite the fact that it has become a very popular medium especially for its affordances (Halpern and Gibbs 2013). Dalia Abdelhady (2011), for example, conducted interviews with Lebanese immigrants living in the diaspora including Montreal, New York, and Paris and tried to find how Lebanese immigrants defined their identities and maintained their relationship with Lebanon. Abdelhady makes a few references to the role of social media in maintaining the connection with Lebanon, saying: Online media provides an important arena where diasporic solidarity as well as consciousness is formed and expressed … Many of my respondents told me that online media allows them to stay in contact with many friends and family members around the world. Exchanges of news, pictures, and conversations were conducted regularly, enabling my respondents to participate in cross-border networks and keep the homeland alive. The ability to share important events as well as routine daily details equally reinforce ties with friends and family in Lebanon or in the diaspora (2011,116).
However, Abdelhady suggests that social media sites do have negative impact especially when some immigrants spend extensive hours each day online which would only isolate them from their new homelands (ibid.). Further, a survey study undertaken by Abdulahad et al. (2009) examined the acculturation patterns of 153 Iraqi Canadians, and only two references were made on media consumption.
Finally, another study which involved researching diasporic media in Ontario examined the use of ethnic websites by several immigrant groups (Karim et al. 2007, 86), however, it did not cover social media channels or Arab Canadians. To sum up, studies on the Arab Canadian community's social media use are lacking, and there is a gap in the literature in this field of research. It is important  to mention here that the online community examined in this study is not representative of Arab Canadians as a whole, and the online community contains members who are not officially Canadian. In the following section, the theoretical framework of the study is presented with a focus on Anderson's 'imagined community' and Barzilai-Nahon's 'networked gatekeeping' concepts.
3 Theoretical background 3.1 Imagined communities and digital diasporas One of the basic aspects of studying online communities is investigating the way various groups negotiate their identity and belonging. For example, Baym (1999) discusses how online audience members form a community that actively interacts with each other. The seemingly strong bond that is established among them ultimately makes them feel like friends which enhances interpersonal relationships as well as their sense of identity. However, this sense of identity and community bond might not always be real but rather an imagined one which is elaborated on by Anderson (1991) who introduced the idea of ''imagined communities''. This concept has been used by hundreds of scholars to refer to online and offline members who are loosely linked to each other and whose idea seem to exist in the minds of the community members alone. This is an imagined community because ''the members of even the smallest nation will never know most of their fellow members, meet them, or even hear of them, yet in the minds of each lives the image of their communion'' (Anderson 1991, 224). However, this ''imagined community'', which is a term applied to communities in general rather than online communities alone, has several positive functions which include offering a sense of unity, hope, and strength to its individuals which will ultimately provide psychological and social empowerment. Another important function is to allow dispersed groups to define their identities and possibly distinguish themselves from other communities or natives in their new place of residence (Cohen 1997). Bonini (2011) adds another function which is providing ''domesticity'' in the sense that media use makes immigrants ''feel at home'' wherever they reside. In other words, Canada itself is regarded as an imagined concept that exists in the minds of people because the definition of this country varies based on different ideological and cultural backgrounds as well as the emotional attachments some people have towards it. In relation to this study, the online community members who gather on the Facebook page examined here, whether they are Canadian citizens or prospective immigrants who plan on moving to Canada, seem to be attached to a subjective notion of what Canada is. Indeed, there is an imaginary connection to a nation that is determined by the formed understanding of the online community members. In this regard, Berry discusses in his model of acculturation (1997) the existence of hybrid or hyphenated identities among many online communities that naturally practice acculturation mechanisms and strategies, especially in the diaspora. Here, the relationship between one's ethnic values and identity on the one hand, and the social involvement with one's host culture, on the other hand, is not a linear or separate process but rather closely interconnected. In other words, many first generation immigrants develop hybrid identities that are shaped by the cultural values of their original homeland and their host nation. Indeed, user-generated content and social media are regarded as one of the most popular platforms in which immigrant groups gather to form their online community and express themselves (Dekker and Engbersen 2014;Lášticová 2014;Marlowe, Bartley and Collins 2016). In relation to digital diasporas, they are organized on the Internet and especially on social networking sites (SNS) (Oiarzabal and Reips 2012), and they can facilitate the ''creation and development of informal and formal migrant networks among individuals, groups, and organizations from the country of origin and the country of settlement'' whose influence reaches the online and offline worlds (Alonso and Oiarzabal 2010, 6). Brinkerhoff argues that digital diasporas exert several types of influences on our societies and international affairs including the ''potential to foster democratic values, support integration in the host society, and contribute to security and socio-economic development in the homelands'' (2009,2). Further, diasporic communities exert a kind of contra-information flow that resists mainstream media channels. This is regarded as globalization-from-below (Karim 2003), involving individuals who share common ethnic or national roots with a genuine need to assert their identities. In fact, people in the diaspora share some ''mythical and linguistic allusions to the ancestral territory, which they invoke in nostalgic reminiscences'', creating a ''demand for cultural products that maintain and ritually celebrate the links of the diaspora with the homeland'' (Karim 2003, 3). This demand for ''cultural products'' certainly involves using SNSs due to their increasing popularity and importance in today's world. Yet, there still remains a significant gap in existing knowledge about SNS use by diasporic communities. Karim stresses that diasporic communities living in the West are among the most active members in producing cultural content. The main motive behind this active engagement in online media is to ''create a virtual community that eliminates the distances that separate them in the real world….Time and space are seemingly held in suspension in this effort to reconstitute the community and to exchange cultural knowledge held in the diaspora'' (2007,373). In this regard, one must wonder how online platforms such as Facebook can provide important venues for horizontal communication? Indeed, online community members often feel empowered to share information and exchange ideas in an equal manner (Tufekci 2014).

Networked gatekeeping on social media
Since the discussion above focused on online communities, it is also important to examine some aspects of online behavior that are linked to the centralized and non-centralized types of communication. This is because understanding the kind of communication flows that take place on social media can shed light on the type of engagement among the members of the online community. When someone creates a public Facebook page, (s)he functions as a centralized figure who can control the information flow (vertical communication), while other users who like/follow the page do not have the same information privilege because they are non-centralized members (horizontal communication). In this way, the Facebook page creator can be a moderator of information which is directly linked to the networked gatekeeping theory introduced by Barzilai-Nahon (2008). This theory exemplifies a form of vertical communication flow. In traditional news media outlets, editors are usually assigned to the task of filtering news production. On social networking sites, normal citizens become gatekeepers by the way they select pieces of information and disseminate them to their friends and followers. This is also applicable to what is known as citizen journalism wherein people are not only gatekeepers but also become part of ''the agenda-setting process'' on SNS which happens ''through active engagement: blogging, reposting, commenting, recommending, rating, tagging and the like'' (Goode 2009(Goode , 1293. In this context, Nisbet mentions the importance of studying social media as ordinary citizens become ''active contributors, creators, commentators, sorters, and archivers of digital news content'' (2010,75). Further, Bruns (2005Bruns ( , 2008 discusses the important role of ''produser'' netizens who function as gatewatchers as they are able to reproduce and filter information on SNS. In their study of Twitter and climate change protests, Segerberg and Bennett (2011) conclude that Twitter Streams ''embed and are embedded in various kinds of gatekeeping processes'' (197), while Bro and Wallberg (2014) stress that we are witnessing today a ''new generation of gatekeepers who use social media to produce, publish and distribute news stories'' (1). In this regard, Shaw (2012) refers to the centralized gatekeeping function in the sense that most of the online discussion is filtered and directed by gatekeepers who use a vertical type of communication that is mostly directed from top to bottom (Castells 2013, 71). However, there is still a decentralized gatekeeping activity performed by the online group in the way they ''filter and moderate each other's participation'' (349), discuss issues, like comments and videos, and share posts. This will be further expanded below in the discussion section. To sum up, the social media outlets that are run by famous celebrities, politicians, and other influentials are marked by a vertical type of communication flow, while their followers practice a horizontal type of communication, which is less effective as they are not in control of the main online posts. However, when there is a collective effort, the followers tend to exert similar or sometimes more influence than the influentials cited above.

Methodology
This study seeks to answer the following research questions: What is the nature of the Arab-Canadian community's online reactions on Facebook? What are the objectives behind and challenges of creating this online community? To investigate these questions, sentiment analysis of a public Facebook page was conducted and the most liked and commented upon Facebook posts were analyzed. Also, data retrieved from Google Trends 1 have been examined to support some of the arguments drawn from the study's findings. Finally, the study involved conducting an interview with the Facebook page administrator to shed light on the reasons behind creating the page and how it is administered.
The public Facebook page analyzed is called ''Arab Canadians'' which was created in March, 2010. In the 'About' section, the page administrator mentions the following in Arabic: ''Canada is our second homeland which embraced us and opened for us all the doors to live with love and dignity. We all love you Canada''. 2 The page had 8627 likes and 3567 people talking about it as of 28 January, 2014. The total number of comments mined was 16,098 and 3887 posts, and the demographic details given based on the self-proclaimed user descriptions indicate that 55.4% (n = 6192) are female, while 44.5% (n = 4977) are males. 3 The Facebook page administrator invited the 1 http://www.google.com/trends/explore. 2 https://www.facebook.com/pages/Arab-Canadians/ 154852551232391. 3 Another Facebook page examined is called ''Arabs in USA and Canada'' (https://ar-ar.facebook.com/Arabs.USA.and.Canada) which was found by making a Google search in Arabic with the search term ''Arab community in Canada''. The page which was created on 14 October, 2011 has had 2404 likes and 81 people talking about it as of 28 January, 2014. In the ''About'' section, the following was mentioned in English: ''For all Arabs in America & Canada''. The total number of comments mined was 20,000. The demographic details given based on the self-proclaimed descriptions reveal that 62.3% (n = 10,720) are males while 37.6% (n = 6466) are females. Yet, it was not included in this study because the content analysis showed that it has been created and visited mostly by a group of Iraqi immigrants in which the personal topics and general issues discussed were irrelevant to this study. researcher to join the Arab Canadians Group following a personal correspondence and a request by the researcher. According to the Facebook page administrator, a great majority in this Group's page are Arab Canadians living in Canada. Based on the self-proclaimed demographics, 52% (n = 23,059) of the posted materials have been shared by females, while 47.9% (n = 21,240) by males. The Group's page, which is closed rather than open, 4 mentions the following in Arabic: ''This page does not have any political or religious affiliation…We are united by respect and our love for our second homeland (Canada)'' (Arab Canadians Group 2014). Similar to the case of the public Facebook page that can be viewed by any online user since it is in the public domain, the textual analysis of the Facebook Group was conducted without any references to names to protect the privacy of its members. In the discussion section, acronyms were given instead of using the actual Facebook user names to further protect the identities of online users.
For the purpose of the study, a webometric tool called NCapture-NVivo 10 was used on 18 January and 8 February 2014 to mine the posts and comments. Gender (male, female) information on the commentators was also collected. Information units include Arabic, English and Latinized Arabic posts uploaded by the Facebook administrator as well as comments by the public. This software is limited because of Facebook API restrictions, so it is not clear whether all the comments and posts that include video links, images, and textual commentary are collected. In addition, background information on Facebook users, aside from gender, is not available. However, one of the limitations of this study is that it did not investigate the 6919 images posted on Facebook; instead, textual analysis has been the main focus of this study.
The collected comments were analyzed using a computerassisted program called QDA Miner 4-Wordstat which is suitable for big data analysis. This study analyzes the text corpus by identifying the most recurrent words and phrases as well as their associated terms that are used in the Facebook posts and comments (Xenos 2008;Park et al. 2011;Al-Rawi 2014, Al-Rawi 2015a, b, 2017b. In some of his studies, Entman used QDA Miner and mentioned that it was useful due to its ''improved efficiency in carrying our more complex content analyses'' (Entman 2010, 334). Further, Miller and Riechert assert that the computer-assisted programs that conduct quantitative content analysis ''allow for an enormous array of precise comparisons and statistical test relating to frame prominence and shifts in frame dominance'' (2001, 119). Matthes and Kohring point out that the advantage of this method is having clear objectivity in locating frames and their associated sentiments (Matthes and Kohring 2008, 261; see also Miller 1997;Miller and Riechert 2001). Since part of this study is focused on understanding the online reactions of the online community, sentiment analysis becomes a suitable method here especially that the data set is large, and it is difficult as well as time consuming to manually code all the Facebook posts and comments. The sentiment analysis approach was used by many scholars cited below who wanted to understand the general sentiments towards different issues such as the online self-representation of the Tea Party Patriots in the USA and the Party for Freedom in the Netherlands (Groshek and Engelbert 2013), citizens' political preferences in Italy and France (Ceron et al. 2014), predictions of general elections (Tumasjan et al. 2010), sentiments towards the U.S. presidency (Hopkins and King 2010), and how newspapers and blogs cover events while expressing attitudes towards certain figures and/or places (Godbole et al. 2007). Yet, the majority of previous studies were confined to Twitter (Go et al. 2009;Pak and Paroubek 2010;Kouloumpis et al. 2011;Agarwal et al. 2011), while sentiment analysis studies involving Arabic or Facebook (see for example Neri et al. 2012; Al-Rawi 2017a) are rare.
In sentiment analysis, it is relevant to investigate how the dominant words and phrases are associated with other expressions because they assist in the overall generation of meaning (Pang and Lee 2008;Taboada et al. 2011). The manual classification of words and terms is done to examine sentiments in different contexts (Diakopoulos and Shamma 2010;O'Connor et al. 2010;Das and Chen 2001;Tong 2001) since it is assumed that ''there are certain words people tend to use to express strong sentiments'' (Pang et al. 2002, 2). In this study, Jaccard's coefficient was used to accurately calculate the associations among the words and phrases by providing a spatial map of the conceptual distances between regularly occurring keywords and phrases. The coefficient ranges between 0.0 for no co-occurrence and 1.0 for complete co-occurrence (Tan et al. 2006;Groshek and Al-Rawi 2013;Al-Rawi 2015b). Finally, the analysis of the Facebook page content through sentiment analysis was complemented by an interview with the Facebook page administrator. The interview was used to shed light on the general purpose of the page, its history, and the way it is being run.

Results and discussion
The results of the study, further discussed below, showed a great number of recurrent words (excluding the prepositions), mentioned by the users who posted and commented (See Table 1 ; Fig. 3). 5 In this regard, Facebook posts refer to text sent by the administrator, while comments refer to the audience messages. Also, the analysis of the Facebook page and its Group page included the examination of the texts accompanying the photos. As for the most dominant phrases used, which are a combination of 2-5 words, we found that the top twenty ones contain Islamic religious supplications and prayers, while two English phrases were among the top ones used which included ''beautiful city'' (n = 18) and ''good morning'' (n = 16).
Facebook posts (n = 3887) focused on different issues related to Canada such as showing and discussing the beauty of Prince Edward Island (rank 3, n = 15), referring to the Toronto Palestine Film Festival (rank 5, n = 11) while religious statements such as prayers were also used (rank 4, n = 15). As for the most active commentators, we find that the user ''Arab Canadians'' who is also the page administrator is the most active one with 683 posts constituting 6% of the total comments.
In the remainder of this section, results are interpreted according to four themes: 1. The online community's social media engagement. 2. The networked gatekeeper and decentralized networked gatekeepers. 3. Facebook and online community.

The online community's social media engagement
To answer the first research question on the social media reactions of the online community, we find from the comments' examination of the most recurrent words and phrases that ''Allah'' is the most cited word (n = 966) which is also evident in the most recurrent phrases used such as ''Allah is glorious'' (Subhan Allah) (n = 142) and ''Allah's willing'' (n = 85). Many of these phrases contain the word 'Allah', but it is not always employed to communicate ideological messages as such phrases are commonly used in daily Arabic language (Piamenta 1979;Edzard 1996;Migdadi et al. 2010 In relation to the above account, the top 20 most liked posts had four instances containing a religious message with a total of 1577 likes, while the remaining ones mostly deal with praising Canada. For example, the top most liked post (n = 472 likes) was written in English and Arabic stating: ''The University of Regina is going to great lengths to help its Muslim students avoid doing the same [sic] when washing for on-campus prayer sessions''. The second most liked post (n = 455 likes) was also written in English and Arabic, stating: ''The first Muslim (hijabi) woman news anchor on American television''.
Aside from the religious terms, we find that the other most recurrent word used is ''love'' (in English) (n = 241). By examining the proximity plot, we notice that this word is firstly and most strongly connected to the word ''Canada'' (coefficient = 0.064) followed by ''city'' (coefficient = 0.043), and ''Toronto'' (coefficient = 0.035) (see Fig. 4). All the other words linked to ''love'' are positive and closely associated with Canada as a whole. For example, the Fig. 4 The proximity plot for the words ''Canada'' and ''Canadian'' in English and their association with other terms Facebook user, F.F., mentions: ''I love Canada even in cold weather…'', while several other users expressed the same sentiments such as ''TA'' and ''HHJ''. The Facebook page administrator himself/herself once posted: ''We all love Canada, the best country in the world!''.
Another dominant phrase used is ''beautiful city'' (n = 18). There were several Canadian regions and cities mentioned in this context such as Toronto, Vancouver, British Columbia, Raincouver, Québec, and Laval city. The Facebook page administrator frequently refers to several Canadian cities such as ''Prince Edward Island'' (n = 15) mostly to describe its features and beauty. Toronto (in English) is also very popular as it is ranked among the top 20 words used by the commentators (n = 90). In relation to Vancouver, R.A. commented in Arabic, saying: ''Vancouver is a city with marvelous innovation and beauty. I can only say here, oh glorious Allah for such a city''. Finally, the word ''amazing'' (in English) which occurred 99 times is strongly linked to various Canadian cities. Here, the online group is mostly expressing a sense of attachment to their present or future host country which is manifested in a more frequent and much stronger way than their link to their original homelands. For example, by examining the word ''miss'', we find that it is mostly used with Canadian cities such as ''Mississauga'' (coefficient = 0.039), ''Montreal'' (coefficient = 0.031), and ''Canada'' (coefficient = 0.018). This could give an indication of the kind of bond created among Arab immigrants with Canada and some of its cities.
Finally, and in relation to the sentiments expressed towards Canada, we find that the country is among the top ten words used by the commentators (n = 202), while the phrase ''love Canada'' is among the first 80 most recurrent phrases. For example, I.H. mentions in Arabic the following: ''Indeed, Canada is one of the most developed countries in the world and Canadians deserve respect and admiration…We hope that Arab governments will respect their people, too''. The third most associated word is ''people'' (coefficient = 0.029). E.A., for example, mentions the following in Arabic: ''Canada is developed not only in its architecture and industry but also in its political system and respect for human rights in comparison to other countries. This is why people immigrate to it to live with dignity''. In the two examples above, as well as in other ones found in the text corpus, Canada is used as a reference to compare living standards and lifestyles to those of the immigrants' original homelands. This online community is typical of Anderson's ''imagined communities'' concept because most of its members have not seen or met each other in real life, but they are linked mostly due to their 'imagined' or perceived shared history and values. Besides, all of the online community members, whether be Canadians or not, have an imagined idea of what Canada stands for, and it seems that this idea, despite all the imagined varieties and impressions that exist, is connecting the online community members together. Since many members of the community are living away from their original homelands, they constitute what is called a digital diaspora (Brinkerhoff 2009), and Facebook provides the technological means as well as the ease to form this online community. Whether they are living inside or outside Canada, Arab immigrants or migrants are using Facebook to connect with each other to exchange valuable information and create important personal networks through Facebook use. Yet, as manifested in the data analysis above, Canada, which is the current or possible future host country, is often used by the online community members as a contrasting example to criticize the ill policies, wrong doings, and mistreatments many users feel towards their original homelands. I argue here that Canada provides a sense of hope and a means of escape from economic hardship, social or religious alienation, and possible political persecution. It is important to mention as well that Canada itself is another imagined community because this idealized vision of the host nation might very possibly change after landing and experiencing the new country.

The networked gatekeeper and decentralized networked gatekeepers
To answer the second research question on the challenges and objectives behind creating the Facebook page, we find that the Facebook administrator leads the online discussion as she sets the agenda for what to talk about (Segerberg and Bennett 2011). The example cited above on posting references and questions on Canadian cities usually generates tens of comments, shares, and likes revolving around the same topic. For example, the most commented on post (n = 94) in English was a simple question addressed to the Facebook page followers: ''Hello form Ontario, where are you from?'', while another one was a question on the recent events in Egypt following the public protests against El-Sisi in August 2013 (n = 54). Also, the Facebook page administrator discusses many other relevant issues such as Canadian universities, cities, living, and events such as ''Happy Canada Day''. This is part of what is called the centralized networked gatekeeping as she directly and indirectly orients the online chatter into a certain way. Further, the interview conducted with her showed that she often moderates the page as ''anything that is not appropriate gets deleted'' (Interview 2014). In other words, the Facebook page administrator functions as a networked gatekeeper (Barzilai-Nahon 2008) who filters comments, posts questions and photos, mediates during debates, and monitors the overall online discussions. This gatekeeping activity, which is practiced in a vertical mode of communication, is part of the agenda setting process on social media due to the Facebook page administrator's active engagement with what the online audience is writing (Goode 2009). By examining the most liked 20 posts, we find that 13 posts deal with praising Canada, the beauty of its cities, and its people which is one way of directing the online comments on Facebook. For example, one post that generated (n = 343 likes) from the Facebook page followers reads: ''At a metro station in Canada, the ticket machines were broken and there were no employees there yet. The result shows just how honest people can be. Awesome''. Another post, which had (n = 259) likes, mentions in Arabic: ''Canada is the third best country in the world in terms of its living standards''. The sentiment analysis provided above indicates that the online community members get directly engaged and actively interact with the kind of questions posted by the page administrator which can explain the highly positive sentiments in these online comments. Further, many online community members are themselves produsers because they do not only passively consume materials posted by the networked gatekeeper, but they also actively produce comments and post images and videos that could provide insight into their values and beliefs (Bruns 2005(Bruns , 2008. Hence, the online community members who liked the Facebook page or are part of the Facebook group can also be regarded as decentralized networked gatekeepers (Bro and Wallberg 2014) unlike the Facebook page administrator who is a centralized gatekeeper (Shaw 2012). This is because they have selectively chosen to like, comment, and possibly share posts from the 'Arab Canadian' Facebook page though they are still limited in their comments to the particular topics suggested by the Facebook page administrator, as explained above.

Facebook and online community
In relation to the purpose of creating the Facebook page, it seems that the administrator wanted ''to inform others about Canada, its news, and charm'' (Interview, 2014). Indeed, the Facebook page functions as a platform wherein the online diasporic community gather, meet others, disseminate and share information, and make important announcements about relevant events. This is also stressed in Abdelhady's study cited above on the diasporic Arab group in North America and Europe (2011). Further, this Facebook page seems to serve other practical purposes. Based on the interview conducted with the Facebook administrator, the majority of the page's followers are indeed from Canada, yet some users who follow the page want to immigrate to Canada and need vital information on visa issues (Interview 2014). This is also confirmed by examining the public Facebook page analytics (Socialbakers 2014). Unlike the case of the Facebook page Group, the analysis of the public Facebook page revealed that most users have originally registered their geographical locations in Canada 36.2% (n = 3121), followed by Egypt 17.6% (n = 1521), and Saudi Arabia 5.4% (n = 464) (Socialbakers 2014). In addition, Google Trends offers another important way to investigate search interest over time and identifying the top cities where a particular search term originates. Using the search term ''Canada'' in Arabic, the top cities that this search term is Googled are as follows: Tripoli (Libya), Damascus, Amman, Riyadh, and Jeddah (See Fig. 5). As expected, the top three most associated terms to ''Canada'' in Arabic are ''immigration'', ''Canada immigration'', and ''immigration to Canada''. The results of the aggregate data retrieved from Google Trends confirm that there is a popular interest in Canada in many Arab cities especially in relation to searching for ways to immigrate to the country. To build up on Anderson's concept of imagined community, I argue that Facebook itself functions as a virtual nation that only exists online and in the minds of people using the platform. Previous empirical studies reached similar conclusions on the way Facebook, and other online venues can function as virtual platforms for churches and mosques (Radde-Antweiler 2008;Campbell 2010;Al-Rawi 2015a, 2017a. In general, the Facebook page followers constitute part of the online diasporic group who believe that they possess a common identity due to their possible shared history, language, culture, and religion (Alonso and Oiarzabal 2010). The members of the 'Arab Canadian' online community, who share a common language and interest in Canada, can enhance their integration into their host countries by exchanging relevant information and news, and sometimes can strengthen their link to their original homelands by reminding them of their possible mutual cultural heritage (Karim 2007). Berry's concept of hyphenated identities (1997) is relevant here since online groups can express sentiments and opinions towards their host country or their original homelands. In some cases, both places are mentioned as a matter of comparison which is clearly manifested in some of the comments cited above. Further, when one examines the top 20 most liked posts, three of these posts contain references to the group's homelands and host cultures at the same time. For example, one photo, which got 311 likes, was posted that contained an adapted line from the famous Tunisian poet, Abu Al-Qassim Al-Shabi, as a funny play of words that reads: ''If the people want to live [in freedom], then they've got to immigrate to Canada…'' The original poem discusses the national struggle for independence rather than immigration. Another post deals with an Arab Canadian university student who designed smart traffic lights and garnered 350 likes, creating a sense of community pride.

Conclusions
This article introduced a study of one online community to gain more insight into the way identities are constructed around virtual nationhood. Social media outlets such as Facebook give the technological affordances for an online community to gather and negotiate what its members think of their current or future country. For the Arab Canadian community, Canada remains an imagined place, but this experience certainly applies to any other online community. In this regard, the Facebook comments and posts indicate that there are dominant religious terms used, yet what is striking in the sentiment analysis of the comments is that most terms are positive toward Canada. This is mostly due to the way the Facebook page administrator asks questions and directs the online discussion, as she acts as a gatekeeper and plays a major role in highlighting the importance of the new or aspiring homeland for Arab immigrants. However, the online community members perform as well a decentralized networked gatekeeping role since they often counter balance the power of the Facebook page administrator, though their influence might not be always strong. Canadian lifestyle, education, and cultural achievements are often contrasted with what is found in Arab countries, and such contrasts seem to strengthen the sense of belonging felt by the online community toward their present or future host country. Indeed, the ''Arab Canadians'' Facebook page functions as a platform to bring together an online community that shares similar values, history, and language. This includes online productions such as literary and artistic content that all play an integral role in linking the online groups to their original homelands and most importantly to their present or future host country. Hence, this is a highly important venue by which this online community expresses views and opinions on different issues. In this regard, Facebook functions as a vital venue for the online community to meet and discuss two 'imagined' homelands: the past/present and the future. This social media outlet also provides the virtual platform for the online community to exchange views, create networks, and most importantly gather important information for possible immigration. In other words, Facebook becomes the online community's virtual and imagined homeland, because this is where the offline community members scattered around the world meets and interact based on their shared language. Indeed, Facebook itself functions as a virtual nation because the platform provides the community members with a much needed sense of identity and belonging to an imagined place that they sometimes long for.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.