Developing Resistance Against Grooming Online - Project DRAGON-S

Individuals who groom children on and offline are also interacting with each other in online paedophile communities where they can trade abusive material, share advice on offending, and try to rationalise their behaviours. These communities can enable offenders to normalise their attraction to children, as well as gain knowledge and skills from other experienced offenders that may lead to an escalation of their criminal activity. Historically, offender-to-offender interaction has been largely ignored in existing child sexual abuse research, remaining understudied to this day. Recently, however, there have been increasing forays into this area by criminologists, psychologists, and now linguists, who are seeking to examine the concerning problem of offenders congregating in online spaces.

Online paedophile communities have been observed in the forms of dark-web forums (Grant and MacLeod, 2020; Woodhams et al., 2021; Huikuri, 2022), public pro-paedophilia advocacy websites (Holt et al., 2020; Nielsen et al., 2020), email exchanges (Luchjenbroers and Aldridge-Waddon, 2011), and common social media sites (Marsh-Rossney 2021; Marsh-Rossney and Lorenzo-Dus, 2022). How these communities are formed and how offenders gain access to them is a question of interest to law enforcement so that prevention measures (like shutting down websites/group-chats and infiltrating communities with undercover officers) can be implemented.

Due to the criminal nature of the interest shared between individuals aspiring to take part in online paedophile communities, members are cautious with newcomers and sharing identifying information. These risk assessment practices are part of the skills that can be learnt by members of the community to build trusting relationships with one another and evade detection or arrest. Once accepted into the community, members trade and share abusive material with each other – sometimes using this media as currency, sometimes using it as a way to ingratiate themselves with their peers and receive praise, or simply for sexual gratification. Another problem with the existence of these communities is the production of an environment that can normalise the sexual abuse of children through offenders encouraging one another and expressing pride in their paedophilic identities. Existing research, including my own work into online offender-to-offender interactions in private social media chats, has found these communities are rife with victim-blaming attitudes and can act as an echo-chamber for offence-supportive beliefs. This all makes it safe to conclude that online paedophile communities facilitate sexual violence against children.

Fortunately, there is a growing effort to investigate and combat these communities by academics, charities, and law enforcement agencies. One such group is the Finland-based NGO Protect Children (Suojellaan Lapsia). This organisation focuses on preventing sexual violence against children through raising awareness, strengthening children’s rights, researching offenders, and rehabilitation methods – and they have conducted vital studies into offender behaviour. Their ongoing ‘ReDirection’ project included a dark and clear-web anonymous survey for individuals seeking out Child Sexual Abuse Material online, which garnered over 20,000 responses and revealed that 46% of respondents claimed to be in contact with other offenders online. In September 2022, Protect Children held an international expert webinar to share their research projects and partnerships (which included Europol, End Violence Against Children, and the charity Stop It Now! amongst others). Speakers included international human rights lawyer Tegan Insoll, who presented on the ReDirection survey results, and Dr Salla Huikuri, from the Police University College (Finland), who discussed their work on dark-web child sexual abuse forums. The latter brought up an aspect of these online communities that my own research is looking into: community roles and hierarchies. Members of these communities can take on roles within the groups/forums like administrators or moderators that manage the content posted. They can also perform identities, like taking on the ‘newbie’ role, acting in an ‘advisor’ role, or that of an ‘expert’ offender. Huikuri claimed in her presentation that targeting users acting as administrators is an effective way of taking down these darknet communities.

The Aston Institute for Forensic Linguistics (AIFL) is another organisation working to combat online paedophile communities, but they are doing so through forensic linguistic research and casework. The institute, directed by Professor Tim Grant, is the largest grouping of forensic linguists in the world and it currently has a strong focus on child sexual abuse research. Work at the AIFL has included collaborative law enforcement projects training undercover officers in infiltrating online paedophile communities by taking on the identities of existing offenders, as well as casework on several prominent criminal cases relating to the takedown of prolific groomers who also operated on dark-web paedophile fora (e.g., Matthew Falder). Academic research in this field is also ongoing at the institute, some of which was presented at the annual British Association for Applied Linguistics Language and New Media SIG seminar held at Aston university in September 2022. The seminar, titled ‘The Anti-Social Network’, included presentations on the ‘manosphere’ and violence on Twitter, alongside talks concerning online paedophile communities. Professor Tim Grant discussed identities, language, and power in dark-web fora – noting how members make ‘claims to power’ by claiming community expertise, technological knowledge, and personal experience for example. Much like the aforementioned community roles, these claims to power can reinforce an offender’s membership position or enable them to ascend the hierarchy to a more respected or powerful position. Dr Ralph Morton and Daniela Schneevogt also presented on dark-web paedophile fora: the former examined the themes present in offender-to-offender interactions, while the latter looked at offender usernames and what trends existed in their contents.

Linguistics has much to offer in the field of child sexual abuse prevention – particularly in the analysis of online paedophile community interactions as this is anonymous written-language data. My own research, for example, explores many different aspects of these communities: from how membership is acquired and maintained to what crimes are divulged or planned out between users. I also look at how offenders construct their sexual identities in these groups and how they construct sexual identities for victims. Much of this feeds into another key phenomenon to examine, that of the normalisation of paedophilia and the supportive nature of these communities that encourage or enable child sexual abuse. Discourse analysis can be applied to data from these community interactions to look at identity construction and the interpersonal dynamics at play in terms of politeness theory, facework, stance analysis, and speech acts. There are many avenues for linguistics to contribute to this field, and an ever-growing need for efforts to combat a problem as consequential as child sexual abuse. Some hope can therefore be inspired by the novel linguistic work being undertaken into offender communities, and by projects like DRAGON-S into online grooming.

References

Grant, T., and MacLeod, N. (2020). Language and Online Identities: The Undercover Policing of Internet Sexual Crime. Cambridge University Press. 26–57.

Holt, K., Liggett, R., Holt, T. J., and Lee, J. R. (2020). Examining Self-Reported Physical Contact With Minors Among Pedophile Support Forum Users. International Journal of Offender Therapy and Comparative Criminology. Vol. 64(4). 299–314.

Huikuri, D. S. (2022). The Darkest Side of the Darknet [ReDirection Report]. Police University College.1–48.

Luchjenbroers, J., and Aldridge-Waddon, M. (2011). Paedophiles and politeness in email communications: Community of practice needs that define face-threat. Journal of Politeness Research. Vol. 7(1). 21–42.

Marsh-Rossney, R. (2021). Constructing the Identity of the Victim and Offender in Online Paedophilia Groups – A Discourse Analysis Approach. Paper Presented at NZ Discourse Conference 8. University of Canterbury, New Zealand.

Marsh-Rossney, R., and Lorenzo-Dus, N. (2022). A Discourse Analysis of Sexual Identity Construction by Offenders in Online Paedophile Communities. Paper Presented at EPICS X. Pablo de Olavide University, Seville (Spain).

Nielsen, M. H., Aaskov, L., and Larsen, J. E. (2020). When virtuous paedophiles meet online: A sociological study of a paedophile community. Sexualities. Vol. 25(5-6). 598–614.

Woodhams, J., Kloess, J. A., Jose, B., and Hamilton-Giachritsis, C. E. (2021). Characteristics and Behaviors of Anonymous Users of Dark Web Platforms Suspected of Child Sexual Offenses. Frontiers in Psychology. Vol. 12. 1–11.

Rosie Marsh-Rossney, Doctoral Student supervised by Prof. Nuria Lorenzo-Dus and funded by the ESRC Wales Doctoral Training Partnership.

From 23^rd to 25^th May 2022, several of us from DRAGON-S attended the tenth International Symposium of Intercultural, Cognitive and Social Pragmatics (EPICS X) in Seville. We were there to find out about the latest research in this area of linguistics, while sharing some of the findings from our own work. It was also an opportunity to spend some time in the charming Andalusian capital, amid the subtropical splendour of the city’s many palms and jacaranda trees in violet bloom. This year, the biennial conference was held at Pablo de Olavide University, where our hosts provided a warm reception and put on a smoothly run event rich in high-quality presentations.

The main focus of our involvement at the conference was an 8-paper panel, organised and convened by our very own Prof. Nuria Lorenzo-Dus (Principal Investigator of DRAGON-S). The title of the panel was “Discourses of and about child sexual exploitation and abuse”, and included papers by speakers from Swansea University, University of Central Lancashire, and University of Valencia; three of the papers were by DRAGON-S members.

The panel brought together a variety of studies highlighting the growing interest within linguistics in this vitally important subject. Papers included several examining the language of children targeted by perpetrators of child sexual exploitation and abuse (CSEA). For example, Mullineux-Morgan and Lorenzo-Dus (2022) looked at the accounts provided by children to helpline counsellors about being sexually groomed online. This paper focused on politeness strategies used in online grooming, as revealed through children’s assessments of groomer behaviour. Pérez-Sabater et al. (2022) also looked at children’s accounts via a helpline, this time in a Spanish context and with a focus on children’s representations of their own behaviour as well as their interpretations of groomers’ behaviour. In this study, such behaviour was analysed using the Online Grooming Discourse Model (Lorenzo-Dus, 2016).

The language of child sexual abuse from the victim’s perspective was also a focus of Fan (2022), which examined accounts from adult males in China who had experienced sexual abuse in childhood. Other research presented on the panel looked at the language of child sex offenders themselves. In Marsh-Rossney and Lorenzo-Dus (2022), the language of perpetrators was examined in the context of online communities of offenders talking to each other. In this study, particular attention was paid to the way child sex offenders discursively construct their own and their victims’ identities.

Overall, this panel at EPICS X showcased the diversity of linguistic approaches to research on CSEA, with papers varying in terms of types of data, topic and the specific methodology used. For example, in addition to research looking at the language of child sex offenders and their targets, the panel included studies that examined legislative texts and transcripts of interviews with probation and offender support staff. The legislative texts were used in Jain (2022) to compare the representation of CSEA in legislation across two national contexts - India and the UK, while Themelidis (2022) used the transcripts to consider representations of female offenders and the implications of these in terms of perceptions of risk.

Other papers on the panel included Mooney (2022), which reported on research that used a mix of qualitative and quantitative methods (discourse analysis and linguistic inquiry and word count) to identify the linguistic features of CSEA where perpetrators attempt to gain direct physical contact with victims; and Evans and Lorenzo-Dus (2022), a methodological paper illustrating how corpus linguistics procedures (keywords and concordances) can be used to investigate online child sexual grooming.

Conferences are great places to exchange ideas, to get to know people with similar research interests, and to gain a sense of how your own research fits within the wider field. At EPICS X, this was true with respect to both the panel’s specific focus on discourse and CSEA and the conference’s broader subject area of pragmatics. The coffee and lunch breaks, with the shade and airiness of their veranda setting, provided a particularly relaxed and convivial space in which to do this.

Beyond the panel, the conference consisted of many papers on interesting and important subjects within pragmatics and discourse. These included papers on relevance theory, politeness and impoliteness, relational work, speech acts (especially apologies and requests), figures of speech, irony, and humour. A number of presentations considered the latest developments and different perspectives of pragmatics as a field of study, and individual papers varied in their focus on topics (from Covid to cancel culture, racism to violence against women) and particular contexts (e.g., the workplace, classroom, online spaces, news media). Several papers looked at discourse and identity, including one from a DRAGON-S member that examined the construction of in-group identities by right-wing extremists (Perkins, 2022).

At EPICS X, there were three plenaries. Prof. Anne Barron gave an engaging talk on speech acts used in first-date payment negotiations and the variation in these based on gender and national cultural contexts. In a fascinating presentation about luxury discourse, Prof. Crispin Thurlow looked at frame-shifting on a TV show about expensive hotels, providing a critical analysis of the ways in which workers are exploited for entertainment purposes and treated as ‘human scenery’. Finally, Prof. Nuria Lorenzo-Dus (DRAGON-S PI) presented a comprehensive and insightful plenary on digital grooming, looking at manipulative discourse in relation to three stances (expertise, avidity, and openness) as adopted in three main types of digital grooming: child sexual, ideological, and (crypto-drug) commercial grooming.

At the start of her plenary, Prof. Lorenzo-Dus decided to break with the convention of first establishing the ‘who, what, where’ of the research, choosing instead to talk about researcher well-being when working with distressing data. The point was to bring to the fore an often-overlooked issue and to challenge some problematic attitudes; for example, the notion that distressing language data is less harmful because ‘it’s just words’, and that coping with such data is simply a matter of being resilient. This challenge will likely have resonated with presenters on the discourse and CSEA panel, as well as other researchers at EPICS X this year. However, very little evidence-based knowledge exists about the impact of distressing research, and it is an area that requires a lot more work. Perhaps this is where academic conferences help bring wider attention to important issues, such as the impact of distressing data on researcher well-being, that urgently need more research.

Dr Craig Evans

In this post, I am going to consider how corpus linguistics as a method for analysing language can be beneficial to research on online grooming communication. By ‘online grooming communication’, I mean the use of language in the context of adults targeting and manipulating children online in order to sexually abuse them. This is an issue that is the focus of our work on DRAGON-S where we are developing tools that will help child-safeguarding practitioners to detect online grooming content and understand groomers’ and children’s communicative behaviour when online grooming is taking place.

Broadly speaking, corpus linguistics refers to a research methodology that uses a corpus or corpora (a body or bodies of texts large enough to mean that manual analysis makes little or no sense) to analyse how language is used. It involves using computer software to identify linguistic patterns – based on frequency or statistical calculations – that can tell us something interesting about a corpus and the use of language it represents; for example, about how minority groups are represented in the press or how people talk about their healthcare experiences online. To do this, a corpus should ideally be representative and balanced. This means it should contain all variations, types of speaker, situations of use, etc, of a language or text type, and in similar proportions to how these occur in the ‘real world’. If it does, then we as researchers can have a certain amount of confidence that our findings are generalisable about the language use a corpus represents.

Being able to produce generalisable findings is particularly valuable when researching the use of language in online grooming. One reason is because the communicative modus operandi of individual groomers varies considerably, from other groomers and within their own communication; such as between being aggressive or friendly, sexually explicit or implicit, opportunistic or gradually tactical, and so on. Therefore, any research that uses small datasets, as often occurs in studies that are purely qualitative, is likely to produce a partial and unrepresentative account of the language of online grooming. Using corpus linguistic methods, on the other hand, means that results can be based on all varieties of this use of language. (Of course, building a balanced and representative corpus using online grooming data comes with its own set of challenges, which is a topic for another day / blog post.)

Another argument in support of corpus linguistics as a method is that it helps reduce the influence of bias. That does not mean it can ever eliminate it completely, and researchers using corpus linguistics have plenty of choices to make where bias will inevitably creep in. However, using quantitative means to generate initial patterns for further analysis does reduce the risk of researchers being overly selective when picking which features or aspects of their corpus to focus on.

Confirmation bias – i.e., when researchers are drawn to samples or features that confirm what they want / expect to find in the data – is a particular risk with research involving well-established ideas about certain practices. This potentially occurs with online grooming, where its abusive and predatory nature may create the expectation that child-targets of grooming are always passive victims of bullying, controlling groomers. However, many child-targets will in fact actively participate in and drive forward conversations with groomers, without realising they are being groomed. It does not make them any less victims, but rather victims of groomers who may be using very subtle means of communicative manipulation.

With corpus linguistics, the use of frequency or statistical measures to generate results not only helps reduce the influence of what the researcher may already think about the data, but it can also reveal patterns not obvious to the naked eye, and lead to new and unexpected discoveries. As I’ve just suggested, groomers can be very subtle in the way they manipulate the children they target. This includes using language in a way that, on the surface, may make them seem like other children, such as when talking about hobbies and interests, relationships with parents, etc. At the same time, though, groomers will be working towards their primary goal of trying to engage children in sexual activity, and this may be evident in underlying patterns in their talk that the use of corpus linguistic methods can help reveal.

Revealing non-obvious meaning or behaviour, and creating opportunities for serendipitous findings, are qualities of corpus linguistics that not only appeal for what they may contribute to knowledge about online grooming, but also for how they help practitioners in child safeguarding roles tackle this problem. Disguise is a hallmark of groomer behaviour, and corpus linguistics has great potential to reveal what groomers may try to hide. In this way, it is a method (or perhaps more accurately, a set of methods) that is well-suited to a project like DRAGON-S and its aim to produce research-evidenced interventions to countering online grooming.

Another benefit of corpus linguistics is that it is very adaptable to mixing methods in different ways. While there are certainly patterns in the talk of groomers that can be highlighted using quantitative measures, online grooming represents two-way interaction between groomers and the children they target. Therefore, it will often involve complex context-specific behaviour that warrants a close qualitative analysis. Corpus linguistics includes a variety of well-established procedures for combining quantitative and qualitative methods, which operate at various levels or dimensions of language – from character to sentence – and always in context. It allows researchers to identify the overall features of how language is used in a particular context, while also being able to zoom in and look at qualitative examples in detail. In this way, researchers are never left unable to see the wood for the trees or the trees for the wood, and can get a complete picture of how people use language.

Thank you for reading this blog post. If you are interested in finding out more about using corpus linguistics to study language use or about how corpus-based methods have been used to research online grooming, here are a few suggestions (by no means a definitive list, just a starting point for anyone new to the topic):

Corpus linguistics and discourse:

Baker, P. (2006). Using corpora in discourse analysis. London: Bloomsbury

Baker, P. and McEnery, T. (2015). Introduction. In P. Baker and T. McEnery (Eds.), Corpora and discourse studies: Integrating discourse and corpora (pp. 1–19). Basingstoke: Palgrave Macmillan

Marchi, A. and Taylor, C. (2018). Introduction: Partiality and reflexivity. In A. Marchi and C. Taylor (Eds.), Corpus approaches to discourse: A critical review (pp. 1–15). London: Routledge.

Partington, A., Duguid, A. and Taylor, C. (2013). Patterns and meanings in discourse theory and practice in corpus-assisted discourse studies (CADS). Amsterdam: John Benjamins.

Corpus approaches to online grooming:

Lorenzo-Dus, Nuria and Anina Kinzel. 2019. ‘So is your mom as cute as you?’: Examining patterns of language use by online sexual groomers, Journal of Corpora and Discourse Studies, 4-39. 

Lorenzo-Dus, N., Kinzel, A., & Di Cristofaro, M. (2020). The communicative modus operandi of online child sexual groomers: Recurring patterns in their language use. Journal of Pragmatics, 155, 15-27.

Lorenzo-Dus, Nuria and Anina Kinzel. 2021: ‘We’ll watch TV and do other stuff’: A corpus-assisted discourse study of vague language use in online child sexual grooming. In Exploring Discourse and Ideology through Corpora, edited by Miguel Fuster-Márquez, et al. Bern: Peter Lang, 189-210. 

Schneevogt, D., Chiang, E., & Grant, T. (2018). Do Perverted Justice chat logs contain examples of overt persuasion and sexual extortion? A research note responding to Chiang and Grant (2017, 2018). Language and Law/Linguagem e Direito, 5(1), 97-102.

Dr Craig Evans is a corpus linguist working as Research Associate within Project DRAGON-S.

“I thought he cared for me” and “it felt like we were in a relationship”. These are some of the comments made by victims of online child sexual grooming when trying to make sense of the abuse they suffered. Indeed, the comments accurately reflect the relationship-building language that online groomers use when interacting with their victims. Phrases such as “im just here to talk” or “I think I’m starting to like you” build victims’ trust as a vehicle to advancing their sexual goals and isolating children, physically and emotionally, from family and other support groups.  

Groomers’ goals are developed in and through language – broadly understood, so words, but also images, videos, etc.  Just how they establish and maintain abusive relationships responds to different stimuli. Situational factors play a part. Digital affordances (e.g., access protocols to different digital platforms), time of contact (e.g., night time), place of contact (e.g., private space in the victims’ home) and so on are all important. Also instrumental are socio-demographic and psychological factors of the groomers and their victims, such as age, gender, mental state and so forth. These variables have been extensively studied within academic disciplines like Psychology and Criminology. Their findings provide an invaluable basis on which to build.

Online child sexual grooming is, first and foremost, a practice of communicative manipulation: one that has to do with the manipulation of trust through language. Profiling groomers’ and victims’ language is crucial to the detection and/or prevention of online child sexual grooming. Linguistics, as an academic discipline, provides a robustly tested toolkit to undertake such profiling. By applying this toolkit, we are able to discern the nuances and hidden meanings of language in use: when – and exactly how -  ‘love’ means ‘sex’, even if no sex-word is used; when offering to help – financial, affective – means threatening to sextort; when seemingly innocent small talk paves the way for addictive engagement online.

Linguistic analysis lies at the heart of Project DRAGON-S. Our effort to find out what language has got to do with online grooming demands we go beyond explicit referents to sexual objectives to uncover also coercive expressions, ‘polite talk’, strategic vague language, and so forth. We then integrate our findings with Artificial Intelligence to develop more effective detection and prevention capabilities to keep our children safe online.

This Infographic shows our research findings into the language of online child sexual groomers.

 Nuria Lorenzo-Dus is a Professor of Linguistics at Swansea University and leads Project DRAGON-S

AI has tremendous capacity to improve everyday lives and combatting online child sexual abuse (OCSA) is among its potential applications. When doing so, of course, it is of paramount importance that AI-based systems are used fairly and transparently. This makes traditional, ‘black-box AI’ approaches unsuitable, as they do not provide users any insights into the internal modelling and analysis processes that underpin them, nor any guarantees that their decision making is correct.

Explainable AI (XAI) is a concept that promises to overcome these limitations. The AI community is working hard at improving its methods, moving towards an XAI whose analysis processes can be understood by non-experts. Once the human decision-maker can readily understand the findings proposed by an XAI system, s/he will be able to assess its levels of relevance and trustworthiness. This will help build user confidence and assure reliable analysis outcomes. 

Project DRAGON-S is contributing to this world-wide effort to create an XAI. Our Spotter tool integrates knowledge from Linguistics into the design of a Deep Learning model to achieve a more explainable and robust analysis. Importantly, it is developed to produce a text analysis system that not only detects and flags possible instances of online grooming in chat conversations, but also provides indicators as to what in these conversations led the XAI system to flag them. When used in law enforcement contexts, the Spotter will therefore triage the conversations that human operators believe should be inspected in detail, and those that they can spend less time on. In doing so, our project will progress both OCSA and synergisation of AIs and human users through XAI. 

Dr Adeline Paiement

Nuria is a Professor of Linguistics and researches digital discourse. Nuria conducts quantitative and qualitative analyses of the communicative (language / visual) tactics used for cyber-crime manipulation. She specialises in the tactics used for online radicalisation and for the sexual grooming of children online. Nuria has an unwavering commitment to putting ethical, innovative and implementable research at the service of collective efforts to keep our children safe online. She loves learning about other cultures and languages, travelling and exercising.

Adeline is an academic researcher in AI and data science, with a physics, computer science, and engineering background. After completing a PhD in computer vision and a postdoc in machine learning, she was a research lecturer at Swansea University for a couple of years, before heading back to France in 2018 to join the University of Toulon. She is particularly interested in multidisciplinary research and in improving AI methods by integrating knowledge of its application domains (e.g. physics, or here linguistics). When she leaves her computer, Adeline enjoys reading and playing music.

Louise is the Project Manager for DRAGON-S and has previously led University projects in global health focussing on burn injury, as well as delivering a European project to grow microalgae using waste. She has worked at Swansea University since 2013, and prior to this Louise managed projects in software development organisations. She studied languages at Swansea University, European Business in Bristol, and is currently undertaking a part time MSc in Psychology. Louise loves camping, trips to the beach and having a lie in when her children permit!

Sophie is the Project Officer for DRAGON-S. Sophie studied Computer Science at Swansea University and went on to complete a Masters in Business Administration. She has held several different roles in project management and operational management in higher education, Welsh Government, 3rd and private sector. Sophie is pleased to be back at Swansea University working on such an exciting project. Sophie spends her free time on 8 wheels ramp skating or gardening with her partner and her 2 little boys.

Craig’s main research focus is corpus linguistics and discourse analysis, and exploring innovative ways of combining the two. His interests include discourse and relational work, the language of persuasion, and the relationship between the personal and the institutional in health and social care contexts. Craig is particularly drawn to research with an applied dimension that can make a positive difference, especially to the lives of socially disadvantaged and vulnerable groups. He also likes films, talking about films, literature and comedy. 

Ruth Mullineux-Morgan’s role within the Dragon-S Project focuses on the development of the Shield Tool for Child Safeguarding Professionals. She is also a final year doctoral researcher in Applied Linguistics at Swansea University, researching children’s discourse in online grooming contexts. Over the last 15 years she has also held various senior policy, public affairs and advisor roles in the political and children’s charity sectors in Wales. Her research interests centre on exploring new ways to promote children’s rights and child protection through linguistic analysis. Ruth is also a volunteer children's counsellor. In her spare time Ruth can normally be found swimming (often in the sea), trying to master stand-up paddle-boarding or curled up with a good book and cup of tea.

Manon is a research associate in Machine Learning / Natural Language Processing, working on developing a chatbot for DRAGON-S  Shield and supporting the development of the project’s Spotter tool. Her previous doctoral research focused on developing multilinguals taggers and parsers for zero-shot models, using typological resources. Outside of work, Manon likes cooking, reading and playing video games.

Chedza is an experienced public relations, media, and government administration and communication management practitioner from Botswana in Southern Africa. He is currently a PhD Commonwealth scholar at Swansea University in the Department of Media and Communication. His research involves organisational crisis communication in Botswana, and investigates the crisis response strategies and the role of social media in disseminating crisis information. Away from the Project Dragon-S and doctoral work, Chedza enjoys watching documentaries and reality shows on television.

Connor Rees is a doctoral researcher in Criminology and Computer Sciences at Swansea University, UK. His research focuses on the ways in which extreme right groups use images similarly to online trolling collectives. Whilst his research covers online extremism, it also covers human-computer interaction and ethics and regulation of Artificial Intelligence within the same remit.

 James is a Computer Science Masters graduate from Swansea University. He focused his studies on machine learning and computer vision, including research into self driving cars and determining where images were taken within the United Kingdom using convolutional neural networks. He has an interest in all things that involve machine learning and artificial intelligence especially if they are used to solve real world problems. Outside of work He likes to spend as much time outside as possible, especially playing sports and walking his dog.

Jay Morgan is a final year PhD Computer Science student at Swansea University. He is working to help build the Artificial Intelligence systems behind the Shield & Spotter tools. Jay's research involves the integration of expert knowledge into Neural Networks to enhance their abilities in performance and robustness, which helps in combining the disciplines of linguistics for the analysis of language in online grooming conversations and the practical usage of Deep Learning for natural language processing.

Keighley is a doctoral researcher in Applied Linguistics at Swansea University, where she is investigating the identity construction of right—wing groups on social media. Prior to this, she completed an MPhil in Applied Linguistics, an MA in Forensic Linguistics, a PGCE (PCET) and a BA in English Studies and Education. Keighley’s research interests include media language, political discourse and health communication. When she is not researching, she enjoys reading, crocheting, cross-stitching and gaming.

Rosie is a doctoral researcher in Applied Linguistics at Swansea University. Her research is in the field of online child sexual abuse/exploitation, with a focus on online offender communities. Rosie has a keen interest in politics, loves reading, and looking after her many houseplants. She is also an avid climber, salsa-dancer, and baker.

Sarah has an MA in Communication, Media Practice and PR from Swansea University. She is currently in the final year of her doctorate into digital literacy education provisions in primary and secondary schools in Wales, a topic she is extremely passionate about. Sarah is also a published author and in a previous life she worked as a PR consultant and a professional vocalist.

Tesni is a final year doctoral researcher in Applied Linguistics at Swansea University. She investigates how Welsh-English bilinguals understand and produce Welsh gender. Away from work, Tesni enjoys playing hockey, cooking and travelling