Social networks are a popular way for people to engage in activities and find new friends over the Internet. One central feature that these services share is the possibility for users to interact and share personal information. However, social networks also enable sexual predators with an interest in children to have access to rich hunting grounds with a reduced risk of detection. This thesis applies grounded theory to provide domain knowledge from eight experts on online child sexual abuse. This domain knowledge is used to create informational structures that can be applied in intelligent software systems to potentially detect these types of crimes. The results provide structures for automated dialogue detection including ontological structures, models of influential circumstances (social and psychological) and a normative dialogue structure of a nested deliberation dialogue. Implementation of intelligent software systems in this domain has the potential to detect probable outcomes of dialogues much faster than human analysts. Future work is needed to provide further theoretical sampling to extend the proposed models and to evaluate the models on real-world data sets. / Sociala nätverk är ett populärt sätt att delta i aktiviteter och hitta nya vänner på internet. En central funktion för dessa tjänster är möjligheten för användare att interagera och utbyta personlig information. Dock medför sociala nätverk också möjligheter för sexualförbrytare att utnyttja barn då de får tillgång till plattformar där många barn befinner sig, samtidigt som de har en minskad risk för detektion. Den här kandidatuppsatsen använder grundad teori för att samla in domänkunskap från åtta experter kring sexuellt utnyttjande av barn på internet. Denna domänkunskap används för att skapa informationsstrukturer som kan tillämpas inom intelligenta mjukvarusystem för att potentiellt upptäcka denna typen av brott. Resultaten bidrar med strukturer för automatiserad dialogdetektion vilket inkluderar ontologiska strukturer, en modell för psykologiska- och sociala faktorer som kan påverka förhållandet mellan offer och förbrytare samt en normativ dialogstruktur baserad på dialogteori. Införandet av dialogmodeller inom intelligenta mjukvarusystem har potential att förutse utfallen av dialoger snabbare än mänskliga granskare. Fortsatt forskning behövs för att bistå modellerna med ytterligare teoretisk applicering samt att utvärdera modellerna på verkliga fall.

<p>The live streaming of child sexual abuse is a serious problem that demands urgent attention from law enforcement, legislature, and the platforms that it occurs on. While child sexual exploitation and abuse crimes have been occurring for decades, the growing number of internet users combined with the recent Covid-19 outbreak has shifted a lot of these crimes online. Crimes that occur over live streams are particularly difficult to investigate. Previous research has highlighted four main types of challenges associated with these types of cases: legislative, social, technical, and investigative. While studies have been conducted that explored these challenges individually, this study fills a research gap by exploring which of these challenges affect law enforcement personnel that work live-streamed child sexual abuse cases post-Covid-19. For this study, an anonymous online survey was sent out to law enforcement individuals and agencies nationwide. The survey was actively distributed for one month. Out of the 29 responses received, 13 participants qualified to take the survey and were asked about the challenges they faced working these types of cases and about any solutions to these challenges. After collecting the data, descriptive statistics identified the challenges, and a thematic analysis was conducted. Following a grounded theory process, the analysis identified five main themes that encompassed multiple types of challenges and solutions. These themes were: a lack of consensus, education, the role of application service providers and platforms, a need for better technology, and out-of-jurisdiction challenges. These themes explored in-depth the participants’ responses and further identified the challenges facing law enforcement that work live-streamed child sexual abuse cases, as well as any current solutions to these challenges. Two main challenges frequently mentioned by the participants were end-to-end encryption and anonymizing services, and one of the main solutions to many of the challenges was the education of parents, children, and legislators. For future research, the author recommends exploration of the following topics: the investigation of the factors that play strong roles in determining the time it takes to investigate live-streamed child sexual abuse cases, research related to how AI could assist law enforcement working these cases, further exploration of a framework for the education of parents and children about social media usage and the risks of live streaming, how social platforms can be used in many ways, and research that examines how often law enforcement encounters live-streamed child sexual abuse cases/whether these cases are common or not. </p>