Trabajo Social y activismo digital: sororidad, interseccionalidad, homofilia y polarización en el movimiento #MeToo

Introduction . Social Work is currently facing the significant challenge of dealing with social networking sites – which have become a parallel universe of socialisation – in which ever-increasing digital activism is taking place. The #MeToo movement stands out as a global benchmark. It has established itself as a digital-global feminist movement, fighting harassment and the abuse of women. Methodology . Adopting a social work perspective, a longitudinal analysis was performed of the #MeToo movement on Twitter between 2018-2019 based on social network analysis and netnography, in conjunction with specific algorithms. Results . The results showed significant patterns of sorority, homophily and affective polarisation through the echo chambers and filter bubbles that were identified in the detected Twitter communities. Furthermore, these online communities reflected real offline characteristics (geographical location, affinities, similarities). Discussion and conclusions .  The #MeToo movement’s global effect and durability has led to a new understanding of social movements in the digital era. Social workers must not be blind to the exciting digital opportunities arising from digitalisation. They must combat homophily and the polarisation of global society on social networking sites, promoting values oriented towards tolerance of diversity. Practitioners must show awareness and intervene proactively in global digital spheres to understand, reflect and promote social justice, equality of rights and the empowerment of disadvantaged, vulnerable and oppressed people.

ever-increasing digital activism is taking place. The #MeToo movement stands out as a global benchmark. It has established itself as a digital-global feminist movement, fighting harassment and the abuse of women. Methodology. Adopting a social work perspective, a longitudinal analysis was performed of the #MeToo movement on Twitter between 2018-2019 based on social network analysis and netnography, in conjunction with specific algorithms. Results. The results showed significant patterns of sorority, homophily and affective polarisation through the echo chambers and filter bubbles that were identified in the detected Twitter communities. Furthermore, these online communities reflected real offline characteristics (geographical location, affinities, similarities). Discussion and conclusions. The #MeToo movement' s global effect and durability has led to a new understanding of social movements in the digital era. Social workers must not be blind to the exciting digital opportunities arising from digitalisation. They must combat homophily and the polarisation of global society on social networking sites, promoting values oriented towards tolerance of diversity. Practitioners must show awareness and intervene proactively in global digital spheres to understand, reflect and promote social justice, equality of rights and the empowerment of disadvantaged, vulnerable and oppressed people.

INTRODUCTION
Addressing the effects of digitalisation -the adoption of ICTs in daily life -is one of various challenges faced by Social Work in the 21st century. It modifies social interactions and consequently, needs, means and professional practices. It is a catalyst for social change, which demands greater and better social justice.
Commitment to promoting social justice lies at the heart of the social work profession (Grant & Austin, 2014). Some define social justice in the terms of oppressor and oppressed language, while others underscore the centrality of promoting human rights, multiculturalism, and diversity (Austin, Branom & King, 2014). The National Association of Social Workers [NASW] (2018) and International Federation of Social Workers [IFSW] (2014) identify the promotion of social justice as a primary social worker role. The Global Agenda for Social Work and Social Development (2020) report, performed by the International Association of Schools of Social Work, the International Council of Social Work and the International Federation of Social Workers, highlighted the role of social workers in co-creating transformative changes through social movements.
Social Work' s emancipatory approach considers activism to be a tool for fighting for better social justice. Research suggests that activists have often used social networking sites as tools of resistance to claim social justice, and shift the balance of power and functioning (Veil, Reno, Freihaut & Oldham, 2015). There is participation in the production and reproduction of digital content, in a process of mass self-communication that challenges conventional structures and promotes the empowerment of those who do (Hoffman, Lutz & Meckel, 2015)

Digital activism, Cyberfeminism and #MeToo
Cyberfeminism is one space where digital activism and social justice intertwine. According to Elbaor (2017) «cyberfeminism is undefined by definition». The term was coined before the expansion of the world wide web. Today, it is used to refer to a social movement which has spread over the internet and uses information and communication technologies. This movement aims at revealing existing gender gaps on the internet, new modes of precariousness and testimonies of domestic violence or lived injustices (Cravens, Whiting & Aamar, 2015). It also constitutes an approach to the potential of certain emerging technologies as spaces of encounter, resistance and vindication against patriarchy (Royal, 2009). Within the framework of cyberfeminism, social networking sites are widely used as platforms for sharing personal experiences (Fawcett & Shrestha, 2016) and from which to participate in public discourse (Cravens et al., 2015). Since 2017, Twitter became a vehicle for an unprecedented wave of revelations of sexual harassments and assaults through the #MeToo hashtag (Sayej, 2017), started by the pioneer, Tarana Burke, in 2006. #MeToo was used 19 million times, with 65% of social networking sites users having encountered content about sexual harassment or assault across platforms (Anderson & Toor, 2018). Viralization on Twitter began when the New York Times published a report detailing allegations of sexual harassment against popular Hollywood tycoon Harvey Weinstein (Kantor & Twohey, 2017). In response, on October 15, 2017, the well-known Hollywood actress Alyssa Milano shared an image on Twitter encouraging women to indicate whether they had ever been sexually harassed or assaulted. The image included the following text: «Suggested by a friend: if all the women who have been sexually harassed or assaulted wrote 'Me too' as a status, we might give people a sense of the magnitude of the problem». Alyssa Milano added a tweet stating: «If you have been sexually harassed or assaulted, write #MeToo in response to this tweet» (Sayej, 2017). Since then, the #MeToo hashtag on social networking sites has been the catalyst for reports of sexual abuse and harassment around the world. It has become a global phenomenon, and an example of active feminism on the internet using social networking sites.

Social Work and digital sorority
Today' s cyberfeminist communicative dynamics transcend geographical and cultural borders, going beyond the realm of close interpersonal relationships and opening the door to so-called «sorority». This word of Greek origin poses a different relational framework to fraternity. In Western culture, the example of fraternity is Cain and Abel, while that of sorority is Antigone and Polynice (Unamuno, 2010). Both examples show the differences between two ways of understanding human relations, however, both are founded on the same principle of Social Work: building relationships to overcome adversities (Addams, 1902). At present, sorority represents an alliance of women who unite around their identity to fight against gender oppression. Varela (2008) understands sorority as the sisterhood of women, which goes beyond differences of class, race, etc. Interactions that can take place through sorority do not have to exist based on lasting or close interpersonal ties, but can occur momentarily, even between unknown women, always based on mutual recognition and respect (Lagarde, 2014). The main function of this type of support relationship is based on survival within the patriarchal system, which in turn is a revolutionary Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo act that breaks the traditional role of the female gender, empowering women and recognizing them as human subjects with the capacity to participate in the first person, so that they can manage their personal and collective lives (Fraser, 2008).
The emergence of this new logic of connective action, based on sharing personalized contents (Bennett & Segerberg, 2012), has reinforced the power of interconnection (Castells, 2012) and has increased sorority potential, giving it a digital dimension.
The increase in social connectivity on social networking sites has made it possible to progressively shorten social distances to an average of 3.5 steps on Facebook (Edunov, Diuk, Filiz, Bhagat & Burke, 2016) and 3.43 on Twitter (Bakhshandeh, Samadi, Azimifar & Schaeffer, 2011), which can reinforce the possibility of a global effect of the fight against the 'patriarchy' and different forms of oppression.

Social work and intersectionality
Social workers and social movements can mobilize intersectionality as a strategy, not merely to include a more diverse range of people, but to examine how we are impacted differently by systems of oppression. The term intersectionality references the critical insight that race, class, gender, sexuality, ethnicity, nation, ability, and age operate; not as unitary, mutually exclusive entities, but as reciprocally constructing phenomena that in turn shape complex social inequalities (Collins, 2015). This literature and current study are founded on the fundamental tenet of intersectionality, i.e., that social identities (e.g., gender and class) are not simply intertwined, but are indivisible from each other and from corresponding social forces (Bowleg, 2012). Cho, Crenshaw & McCall (2013) view intersectionality as an analytical sensibility, arguing «what makes an analysis intersectional is its adoption of an intersectional way of thinking about the problem of sameness and difference and its relation to power» (p. 795).
Examining trends in digital communication patterns -which have been produced within the rubric of intersectionality as an analytical strategy -may be more productive, together with trying to determine how interpretive communities within the #MeToo social movement set the stage for intersectional analysis.

Homophily and affective polarization on social networking sites
The current congregating power of social networking sites around certain political demands has provoked online disinhibition (Suler, 2004), thereby losing fear of positioning itself in the face of normative social expectations. Not only are users no longer hidden -their visibility has become a liberating factor that implies an increase in reputation. Therefore, different actors converge and unite in real time to «do public things» (Latour, 2005) and claim voice and recognition (Couldry, 2004) on social networking sites. The digital context institutes practices that are adopted and reinterpreted by the feminist subject. In this digital medium, a whole multitude of online connections are intertwined around affinities -including gender -which form online communities that are invisible to the naked eye. These communities are shaped around our tendency to surround ourselves with those who share our perspectives and opinions. However, when it comes to politics or culture, homophily can amplify a tribal mentality. McPherson, Smith-Lovin & Cook (2001) state «homophily limits people' s social worlds in a way that has powerful implications for the information they receive, the attitudes they form, and the interactions they experience» (p. 415). Homophily is the basis of the ideological echo chamber effect, that is, the tendency of people to communicate with those who share their political views, thereby creating homogeneous groups. Sunstein (2009) highlighted how «echo chambers» emerge in online environments, where people choose to preferentially connect with each other, to the exclusion of outsiders (Bruns, 2017). Echo chambers are the underlying network structures that provoke the so-called filter bubble. Pariser (2011) pointed out how a filter bubble emerges when a group of participants, independent of the underlying network structures of their connections with others, choose to preferentially communicate with each other, to the exclusion of outsiders (Bruns, 2017). Both echo chambers and filter bubbles degrade the quality, security and diversity of online discourse and influence their subsequent beliefs and actions (Gillani, Yuan, Saveski, Vosoughi & Roy, 2018).
These inertias, among others, have contributed to increased levels of affective polarization, that is, the strong negative emotions that members of one particular party feel for those of another (Iyengar, Sood & Lelkes, 2012). This affective polarization is particularly worrying for the quality and nature of civic discourse, as it is often based on tribal loyalties; forming different groups which have the same ideological beliefs, which avoid debate and rational discussion on issues outside of said beliefs. This lack of tolerance in online networks leads to mirrored behaviours in offline spaces. Several empirical studies have Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo attempted to better understand how social networking services can exacerbate polarization. Yardi and Boyd (2010) showed how responses among like-minded Twitter users occur more often than among users who differ in their political views, and that discussing a highly politicized topic tends to strengthen the group' s identity.
In terms of the different dimensions of the digital world, the observation of social networking sites can be seen as a data collection and social analysis environment. It acquires a central meaning when understanding the communicative and political processes in the cyberfeminist sphere, and it is also a place for applying Social Work perspectives. This is such because in these environments, relationships of exchange, cooperation, conflict and deliberation are established, in which identities, meanings and imaginaries are generated and negotiated. A good 'place' for performing this type of analysis is the so called #MeToo movement.
Starting from the hypothesis that communicative action forms part and is, in turn, constitutive of the containment policies of these movements, the hypotheses that guided this research were as follows: 1. People interact and create communities of different affinities in the #MeToo conversation on Twitter. 2. Most communities emerging around #MeToo share the same beliefs, thoughts and political views, contrary to other communities, reflecting affective polarization on Twitter. 3. Interactions around #MeToo claims also reflect other aspects of discrimination (race, ethnicity, etc…), which go beyond abuse and sexual harassment. 4. Emotions are a catalyst for conversation and testimonies around abuse and harassment.
The ultimate research objective is to detect communities, analyse cohesion and identify leaders from the conversation on social networking sites, as «communication repertoires» (Mattoni, 2013) and «collective action», and interpret whether it has an equivalence with certain offline affinities. If we return to the essence of Social Work, the fight against oppression and for social justice must be a priority. In this case we analyse, from a social work perspective, different digital activism events against sexual abuse and harassment to understand what the social demands are, how a global discourse is formed and articulated globally on social networking sites.

METHODOLOGY
Although Big Social Data provides samples of massive data around social networking sites and allows us to analyse millions of people in real time, it has a significant drawback, i.e., explaining why users of these services do what they do (Boyd & Crawford, 2012), thereby making it difficult to interpret their Joaquín Castillo de Mesa, Chaime Marcuello-Servós, Antonio López Peláez y Paula Méndez Domínguez behaviours and attitudes. To fill this gap, Wang (2013) states that Thick Data is needed, a term that points to Geertz' s (1993) concept of «dense description» as a method of analysing phenomena, cultures and relationships between people. Thick Data offers more valuable information, more personal data, from a smaller sample, which allows you to contextualize and interpret these data. Twitter hashtags allow that. 'Hashtags' were developed to organize information and discussions, but they have become meta-discourses, through which people give context, emotions, and meanings to their posts, as well as find others gathered to share and discuss the same issues (Suk et al., 2019), where people who use them develop feelings of connectedness and ties through a shared language, emotions, and experiences (Dixon, 2014).
To find this context from which to analyse digital activism around oppression and social justice, the Hashtagify tool was used to identify the most popular hashtag related to sexual harassment and abuse, finding that the hashtag #MeToo catalysed the conversation and global interactions on Twitter about this event.
The Gephi software, version 0.9.2, (Bastian, Heymann & Jacomy, 2009) and the Twitter Streaming Importer plugin were used to extract Twitter data. In order to access the Twitter data, it is necessary to enter the developer credentials. Once the credentials have been entered, the monitoring parameters are configured, including: users to follow, keywords and type of network. In the extraction and processing of data from Twitter, communication and interaction with those observed was avoided in order to rigorously comply with ethical criteria. The confidentiality and anonymity of participants in the presentation of results was not necessary, as all the data is public. Once the data were downloaded, a systematization and classification process was carried out to facilitate the analysis.
We have studied the #MeToo movement between 2018 and 2019 in order to find this Thick Data -from which we can understand and interpret an environment within Big Social Data. The #MeToo movement was chosen because it was ideal for following the thread of a conversation on a global scale, which would give meaning to the results obtained. The evolution of social cohesion and the comparison of detected communities and leaderships can be analysed using the conversation around the #MeToo hashtag on Twitter. Four random samples were collected in a longitudinal analysis at four points in time, between September 2018 and March 2019. Parallel research processes were undertaken in order to verify these hypotheses. Firstly, different samples were taken, and the Force Atlas 2 distribution algorithm was applied. It simulates a physical system in order to spatialize a network. Nodes repulse each other like charged particles, while edges attract their nodes, like springs. These forces create a movement that converges to a balanced state. This final configuration is expected to help with the interpretation of the data (Jacomy, Venturini, Heymann, & Bastian, 2014). Secondly, social network analysis allowed us to measure different relational properties, which show the characteristics of the online social structure. The degree centrality is conceived as the number of actors to which an actor is directly linked (Brandes, 2001). Another of the relational properties analysed was the 'betweenness centrality' of the network nodes. This measure calculates the frequency with which a node appears on the shortest path between the nodes of the network (Brandes, 2001). Both metrics define the leadership and capacity of influence. Cohesion analysis is used to consider the measure of closeness centrality, which is defined as the mean distance from one node to all other nodes in the network. Closeness centrality emphasizes the mean distance from one actor to others by focusing on the geodesic distance (Brandes, 2001), i.e., the shortest route a node must follow to reach other actors in the network. The higher average closeness metric, the lower the average level of proximity of the nodes to the rest of the nodes. Cohesion levels were compared based on the samples collected. Specifically, the small world effect was reformulated using Latapy' s clustering coefficient algorithm (2008), which defines the clustering coefficient of a V node as the probability that any pair of randomly chosen nodes are neighbours of V and are linked to each other. The average distance to each other in the four observed samples was compared. Thirdly, communities were detected using the modularity algorithm (Girvan & Newman, 2002). This algorithm is a method of community detection. It consists of decomposing the analysed social structure into communities. Using social network analysis methodology as a springboard, the technique of statistical modularity was employed. This is a method of community detection that consists of decomposing the analysed structure within communities. This Joaquín Castillo de Mesa, Chaime Marcuello-Servós, Antonio López Peláez y Paula Méndez Domínguez algorithm begins by considering all the isolated nodes to later determine if ties can be found within a community or between the community and the rest of the network. In effect, it follows an accumulative strategy. Removing the nodes with the highest level of betweenness causes the communities to become more segmented and defined. Conglomerates are successively formed based on a higher increase of modularity. The process is suspended when the maximum modularity possible between pairs is reached. The way this statistical technique optimizes the division of communities is what makes it more empirically reliable. It makes an adjustment according to degree centrality, i.e., according to the possibility that a tie between two nodes exists, which is proportional to its degree. All these social network and algorithms metrics were obtained using Gephi (Bastian et al., 2009). Fourthly, in order to determine the factors that shaped the communities detected on Twitter, we used netnography or digital ethnography, which can be defined as a set of methods for recording and interpreting digital environments that attempts to adapt the notions and guides of classical ethnography to new places of technological mediation (Hine, 2005;Kozinets, 2015). The principles of netnography are based on the continued immersion of the 'netnographer' in a place of communicative interaction (Hine, 2005), in this case, the #MeToo movement on Twitter. This 'immersion' is not only about accessing and distinguishing the content of information shared on Twitter, but is also an introspective, reflexive process, within which the culture of the medium and meaning of symbols derived from subjects' behavioural patterns in this digital context are also analysed and interpreted. In netnographic analysis, not only is text dealt with, but also the interpretation of images, photographs, videos and other digital appliances. Moreover, netnography is linked to semiotic research on social networking sites as many key images and signs are used that must be interpreted to understand the intersections of language (Bartle, Avineri & Chatterjee, 2013). To better understand the object of analysis, notes are taken on forms of communication (Emerson, Fretz & Shaw, 2011) and screenshots are made of initial impressions about communities, key events, contributions and perceived contingencies in relation to observed users. To address possible validity and reliability biases in the interpretation of interactions, the analysis was conducted simultaneously by three different people, discussing the interpretation and reaching consensus on the meaning of the content of the interactions.
Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo

RESULTS
The results relating to the three fundamental sections of this research: social network analysis, detection of online communities and netnography, are set out below.

Social Network Analysis
The results of the relational properties analysed defined the morphologies of the #MeToo networks observed. The average degree centrality in all samples has evolved from 2.03 to 2.354. This metric indicates the average number of contacts that any node is connected to. As we can observe the amount average of contacts are very low. This low level of average contacts shows that the majority of relationships are weak, that means the close contacts were not predominant. Social distance has shown the average distance between the nodes in the samples observed. The average distance in the four samples showed hardly any differences. We can see how nodes with levels that ranged from 1.632 to 2.309 stood out in the samples, which is under the theoretical average value evidenced on Twitter: 3,43, indicating very optimal proximity. At same time, closeness values are relatively high. The mean value of the clustering coefficient varied in the three samples collected from 0.69 to 0.89. This value, which can be scaled between 0 and 1, reached an optimal level in all four samples.

Detection of communities
Through the statistical technique of modularity, which allows the identification of dense clusters of relationships in broad social networks (Girvan & Newman, 2002), a division of the social structure analysed in 'communities' for each sample was obtained.  As shown in Table 3, the corresponding modularity measures are reflected in each sample analysed.
These communities are implicit: a description resulting from an external observation, based on the data. These detected communities are differentiated by colours in the resulting figures (1, 2, 3 and 4). The modularity values of the analysed samples oscillated between 0.5 and 0.7, being considered optimal values since the adequate parameter of this measure must oscillate between 0.3 and 0.7 (Girvan & Newman, 2002). Figure 1 shows that among the 298 communities detected, 5 stood out, which amounted to 38%. The yellow community was the largest, with 14.17%. This was followed by green (5.84%), light blue (5.47%), blue (5.14%), orange (3.77%) and red (3.63%). Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo In Figure 2 the red community represented 26%. The blue community also stood out with 22%, the yellow community had 18%, the green community 16% and the orange community 11%.
In Figure 3, of the 114 communities detected, there were 4 that accounted for 82% of the nodes participating in the conversation. The yellow community stood out with 46.97% with the green and light blue communities amounting to 12.7% and 12.2% respectively.
Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo

Netnography
Once we detected the communities, we analysed what was being talked about in each one.
In the yellow community in Figure 1, there was a conversation about the case of an Iranian woman imprisoned for uploading a dance video to YouTube. The reaction of the rest of the women, led by an Iranian activist, was to upload videos with them dancing onto YouTube in solidarity. The light blue community conversation in Figure 1 revolved around the case of the salary gap at Google, and how all workers stopped work in its workplaces all over the world to condemn the situation. The red community talked about a women' s demonstration in Kosovo over a case of abuse. The green community focussed on the sexual abuse of actresses in Hollywood and the changes resulting from this. The dark green community discussed the appropriateness of Alyssa Milano positioning herself in favour of a Muslim activist. Finally, the fuchsia community discussed the role of the women' s march in the USA mid-term elections.
In Figure 2, the light blue community discussed politician Joe Biden's alleged abuse. The dark blue community discussed the case of a famous singer who purportedly used her femininity to manipulate, drug and rob men. The fuchsia community discussed cases of sexual abuse at the Autonomous University of Nuevo León. In the green community, there was debate about the possible construction of a wall between Mexico and the USA, and the position of both countries on migration policy. The orange community discussed the relationship between #MeToo and a decrease in sexual desire.
In Figure 3, in the yellow community, a prominent Republican commentator, Stephanie Hamill, slipped in the idea of whether a certain way in which women dress may cause increased instances of sexual abuse. Faced with this view, a wave of women in the green community expressed indignation that no form of dress implies sexual consent. In the blue community, there was debate about whether certain forms of dancing, such as twerking, are too suggestive, and whether this is the gender identity that the women of #MeToo are looking for. In the orange community, following a news piece that highlighted cases of abuse of women in London, there was debate about the victim' s guilt.
In Figure 4, the yellow community' s conversation focussed on the case of actress Thelma Fardin, and an accusation of sexual abuse by her former co-worker, actor Juan Darthés. Additionally, it derided the hypocrisy of certain actresses who now joined in, but had not spoken out before. The green community focussed on the Megan Fox case and her concerns about publicly making accusations about sexual abuse committed by Harvey Weinstein, because she did not believe she had the approval of others. In the fuchsia community Joaquín Castillo de Mesa, Chaime Marcuello-Servós, Antonio López Peláez y Paula Méndez Domínguez (3.49%), the conversation focused on the content of a tweet which condemned the verbal sexual harassment suffered by a girl on the subway, attaching a video for visual reference. In said video, we can hear the incessant comments directed at the author of the video, such as «vous êtes magnifique», together with lingering stares directed at her. The girl explains in her own tweet that this is what she suffers daily, and mentions the customer service account of the urban transport company, RATP, hoping that action will be taken against this behaviour. Furthermore, this tweet reached 36590 retweets, indicating that it achieved broad diffusion. In terms of the blue community, we found the diffusion of a tweet that condemned the fact that people like Soros or organizations like the United Nations are behind the #MeToo movement; that is to say, well-known figures with power. This could indicate that there are certain conspiracy cognitions, affirming that the objective of #MeToo is to control the population. An in-depth analysis of the affinity focuses that convey interactions in the different communities of the samples observed showed the following were prevalent: geographic space/language, common interest, political ideology, popularity or fame of the 'twitter actors'. When the content of the interactions in each of the communities has been analysed, we have found that there were very high levels of similar content whose affinity was structured by these different characteristics. In some communities the language was the factor that defined affinity, which used to be related to a specific country when the different posts (tweets, mentions, hashtags, media, links) were interpreted. In this context, the influence of certain popular leaders caused interactions to revolve around their posts, which reflected their ideology. In all samples observed, and in most of the communities detected, strongly marked thematic ratios of common interest were reached, ranging from 78% to 100%, with a degree of affinity found on common themes in each of these communities. The conversation almost always originates with condemnation of a case of abuse or sexual harassment, followed by a thread of retweets, only sometimes with comments; flooding the community with comments of the same nature, leading to a pattern of homophilic behaviour. These are the densest parts of the figures. It is relevant that, in the same sample, another community may quite often appear with another thread of retweets that criticizes or questions the origin or cause of the complaint of abuse or harassment. Habitually, the original tweet creates a debate which causes doubt and opposition; a heavily critical thread of conversation is generated. For instance, communities have been created, like the blue one in Figure 2, around the condemnation of alleged abuse by politician Joe Biden. Faced with these cases, communities appear Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo simultaneously -for instance, the yellow community in Figure 2 -which call into question the allegations of related abuses. Also, the #MeToo (red community, Figure 2) movement is itself called into question. Therefore, political alignment determines the conversation on social networking sites, and from this alienation, affective or ideological polarization has become evident.
Public figures (politicians, famous actors and actresses, etc.) have different patterns of leadership and influence. Once they share a tweet, either condemning an action, or in solidarity with an instance of abuse or harassment, this immediately generates a thread of retweets that give rise to communities. Quite often opposition to a position taken by these public figures generates much more interaction. This is true of the green community in Figure 1, where the veracity of Alyssa Milano, a popular actress and activist, is questioned for having positioned herself in favour of a Muslim activist. It must be said that the film industry has had a highly active online presence in terms of Hollywood accusations. For example, in the green community in Figure 1, patterns of solidarity have also been found. When there was a flagrant case of penalization or abuse of women' s rights, a wave of demands in the form of tweets and retweets appeared supporting the causes.
There have even been solidarity movements for the victim. This is true of the yellow community in Figure 1, where the case of the Iranian woman recorded dancing on YouTube and subsequently imprisoned resulted in a wave of women acting in solidarity with the victim, and videos of them dancing poured onto YouTube, which were also shared on Twitter with the #MeToo hashtag. Emotivity, therefore, is a feature of high participation. It is also possible to see how communities which claim women' s rights in relation to the wage gap are formed. The case of the light blue community in Figure 1 is significant. All this community' s tweets and retweets are related to the strike action by Google workers all over the world.
Affinity was also identified according to geographical area: for example, the cases of abuse in Iran (yellow community, Figure 1), Kosovo (red community, Figure 1), Argentina (yellow community, Figure 4), Germany, etc. The case of the green community in Figure 2 is significant because it unites gender denunciation with xenophobic accusations, as is the case of the threat to build a wall in Mexico.

DISCUSSION
Analysis of the #MeToo movement shows a social phenomenon of global dimensions. Sorority is an alliance of women which has strongly reproduced in the digital sphere, generating global digital solidarity which has encouraged Joaquín Castillo de Mesa, Chaime Marcuello-Servós, Antonio López Peláez y Paula Méndez Domínguez participation, in such a way that online 'disinhibition' has expanded. A type of 'disinhibition' in which the conversation descends to the most irrational, primitive thoughts that provoke the indignation and vindication of rights, status and responsibilities of today' s women without fear; even with a possibility of improved reputation, drawing on emotions and outrage to create meaning and mobilize related movements (Bhyuan, 2018). Social movements used to have a life cycle, which reached a peak during the first stage of demonstrations and then descended progressively, to then be diluted. However, the #MeToo movement has been analysed longitudinally and has demonstrated durability, maintaining its ordering effects on social networking sites through time (Coenen et al., 2012).
We have analysed the connectivity and interaction around the #MeToo. The metrics obtained reveal that the network structures of the collected samples are highly interwoven, since the four samples reach average distances below 3.43, the average distance on Twitter. It demonstrates that different communities of participants chose to preferentially connect with each other, to the exclusion of outsiders, forming echo chambers. The more connections were created within the communities, and the more connections with outsiders were severed, the more isolated from the introduction of outside views is the community; while the views of its members can circulate widely within it. According to the underlying network structures of their connections with others, participants communicated with each other regarding a determined topic, to the exclusion of outsiders, creating filter bubbles. The more consistently they adhere to such interactions, the more likely it is that participants' own views and information will circulate amongst group members, rather than information introduced from the outside.
Some intersections with other social problems -such as ethnicity or race -are being reflected in global movements like #MeToo. We have found important connections among the core ideas of community organizing, identity and interlocking oppressions. The #MeToo movement illustrates the links between social movement and community organizing sensibilities, the claim that intersectional frameworks were needed to address the social problem of violence against women of different ethnicities, and the call for identity politics to empower coloured women.
In terms of intersectionality, a typical criticism highlights the #MeToo movement' s narrow focus on cis-gendered issues in the workplace -especially accountability for perpetrators and policy change within professional settings. However, concerning the Google worker demonstrations, reflected in #MeToo, we have detected how this movement can remove the gender question on the Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo offline workplace, demonstrating how SNS often become a space where citizens gather and contest the definition of a movement, potentially translating this energy into offline actions. The same interests and affinities or common language project the effect of a social mirror -that is -one that demonstrates that online connectivity and interaction are no more than a faithful reflection of socialization in offline reality (Dunbar, Arnaboldi, Conti, & Passarella, 2015). Therefore, intersectionality is simultaneously reflected. Additionally, it is mirrored around ideological and political affinities. The detected communities have been formed around interactions on equal beliefs, giving rise to very intense patterns of homophily. This leads to polarisation, most often catalysed by affectivity (Iyengar et al., 2012). The increase of «infoxication and infosaturation» of content on social networking sites becomes manipulation and conflict -in the cyberfeminism realm as well -in which patterns of conduct that seek to harm the cause by belittling the #MeToo movement are manifested and reinforced, diminishing the empowerment of women through these global movements.
Personal testimonies and expressions have a mobilizing power, especially through visibility and accessibility achieved by frequent sharing on social networking sites, via the networked acknowledgement that they are taking control in this way (Suk et al., 2019). It is an action which is aimed at eliminating oppression in the social and political environment, therefore, an action of the people. This collective action is directed towards a desire for socio-economic, political and/or structural changes that will bring about a more just balance of power. Women have been empowered by #MeToo through the conjunction of social action, political awareness and the right to share their experiences of harassment and abuse.
Empowerment is based on a process of liberation from 'voicelessness' or silence, which involves raising awareness (Breton, 1994). To enhance this empowerment, such action must be accompanied by reflection, or involve what Freire (1970) calls 'praxis', that is, a constant movement from reflection to action and back to reflection. One must learn how to use one' s voice to achieve social change and to obtain resources for it. The process of empowerment implies competence: it implies ensuring that the right to speak becomes synonymous with the right to be heard. These testimonies must prevail in spaces of diversity for the empowerment process to be operative, building coalitions across diverse identity positions, creating appeals that resonate with a wide swathe of citizens, generating sustained engagement in political action, and supporting intersectional voices in feminist movements and social work issues.

CONCLUSIONS
Digital activism is breaking boundaries through #MeToo, empowering feminism and constituting a transnational sorority. Women have found a complementary medium to catalyse complaints about cases of harassment, assault and abuse and gain support among themselves to vindicate oppression throughout the world. However, the #MeToo movement faces a particular definitional dilemma -it participates in the very power relations that it addresses and, as a result, must pay special attention to the conditions that make its knowledge claims comprehensible.
Although online communication facilitates these new social movements, which have experienced a great boom and diffusion, we are also hearing voices that warn of growing manipulation around these forms of interaction and socialization (Lanier, 2018). There are myriad ways in which feminist practice and praxis are manifest in social work (Park, Wahab & Bhuyan, 2017). As social workers and scientists, we must pay attention to and investigate these means of digital activism in order to understand and avoid possible manipulation and disinformation. From a Social Work point of view, it is possible to contribute to awareness and critical reflection of what this means. The implications for social work practice now are not only to adopt and use technological tools, but to improve digital skills that allow the extension of the capacity of digital diagnosis, awareness and pedagogy to promote digital empowerment from a social work approach.
As social workers, we must abandon the role of expert observers, learn from these movements and what role can individual social workers or the profession as a whole play in bringing out the change this indignation demands. We need to learn how to use digital media as a vehicle for social justice expression, demand and share data, information and digital content with reliability, know how to interact through digital technology using criteria to protect personal privacy in digital environments using the appropriate means, according to the objective.
From a social work practice approach, we can act as a radar for reliable information, detect false information early and turn reliable information into narratives that raise awareness in society. Social workers can become sources of micro-information, facilitating the creation of digital content in different formats from a gender perspective, and the appropriate creation of networks for participation and collaborative work, as well as promote strategic alliances. It is possible to use social networking sites from a digital Social Work approach to building spaces for socialization, with Twitter, LinkedIn and Facebook among the major actors in Social Work -even amongst clients/users. Network Social Work and Digital Activism: Sorority, Intersectionality, Homophily and Polarisation in #MeToo intervention strategies to promote sorority and combat deliberate fakes can be designed in these spaces; the strategic use of digital activism as a tool for strengthening networks of women and other excluded groups; promoting participation of women in today' s digital society and seeking opportunities for self-empowerment and participatory citizenship through the appropriate tools, participating in the design of applications and technological tools.
Finally, regarding social work education, Bhyuan (2018) invoked people to «take seriously the responsibility to bring social movements into the classroom -to examine current social problems and to encourage students to connect their social work practice to processes that bring about transformative change» (p.1). As social work educators, we need to train future leaders who engage meaningfully in digital activism with the individuals and communities with whom they work, capable of attending an urgent need to expand digital skills and build capacity for the most vulnerable people to manage digital identity, fakes and avoidance of manipulation or provocation on social networking sites. In short, Social Work must take up the challenge of digital transformation.

FOUNDING SOURCES
The English translation of the article was funded by the Faculty of Social and Labor Sciences, University of Malaga, Spain.