Publications

The last decade produced rapid developments and powerful new technologies that are creating a huge upsurge in artificial intelligence research. However, for critical operational decisions (e.g., consulting services), the need for explanations and interpretable results are becoming a necessity. The integration of knowledge graphs that provide relevant background knowledge in machine-readable form, and machine learning methods represents a new form of hybrid intelligent systems that benefit from each other's strengths. Our research aims at an explainable system with a specific knowledge graph architecture that can generate human-understandable results even when no suitable domain experts are available. Against this background, the interpretability of a knowledge graph-based explainable artificial intelligence approach for business process analysis is focused. We design a framework of interpretation, and show how interpretable models are generated. Result paths on weaknesses and improvement measures related to a business process are used to produce stochastic decision trees, which improve the interpretability of results. This can lead to interesting consulting self-services for clients or be applied as a device for accelerating classical consulting projects.

The depiction of gender is the focus of a growing number of content analyses in the fields of both mass media (e.g., Goffman, 1979; Grau & Zotos, 2016; Mitchell & McKinnon, 2019; Sink & Mastro, 2017; Ward & Grower, 2020) and social media (e.g., Baker & Walsh, 2018; Döring, 2019; Döring & Mohseni, 2019; Döring et al., 2016). Typically, the depiction of gender follows traditional gender roles and, hence, does not include at lot of individuality and diversity but sticks to established gender stereotypes (Collins, 2011). Gender steoreotypes are defined as beliefs about how men versus women are (descriptive beliefs) or should be (prescriptive beliefs). Relevant dimensions of gender stereotyping are occupations (e.g., the man as the hero, breadwinner, or executive; the woman as the mother, housewife, or subordinate), sexual and romantic behaviors (e.g., the man seeking sex; the woman seeking love), personality traits (e.g., the man being active, aggressive, rational, and instrumental; the woman being passive, affectionate, emotional, and social), or body types (e.g., the man being tall, muscular and older; the woman being petite, slim, and younger). Gender stereotypes in the media cover different dimensions of traditional masculinity and feminity and are represented textually and/or (audio-)visually. Typically, the occurrence and nature of gender stereotyping in different media is measured and changes over time are of particular interest (e.g., Bhatia & Bhatia, 2020; Maker & Childs, 2003).

In recent years, various depth sensors that are small enough to be used with mobile hardware have been introduced. They provide important information for use cases like 3D reconstruction or in the context of augmented reality where tracking and camera data alone would be insufficient. However, depth sensors may not always be available due to hardware limitations or when simulating augmented reality applications for prototyping purposes. In these cases, different approaches like stereo matching or depth estimation using neural networks may provide a viable alternative. In this paper, we therefore explore the imitation of depth sensors using deep neural networks. For this, we use a state-of-the-art network for depth estimation and adapt it in order to mimic a Structure Sensor as well as an iPad LiDAR sensor. We evaluate the network which was pre-trained on NYU V2 directly as well as several variations where transfer learning is applied in order to adapt the network to different depth sensors while using various data preprocessing and augmentation techniques. We show that a transfer learning approach together with appropriate data processing can enable an accurate modeling of the respective depth sensors.

Pornography is a fictional media genre that depicts sexual fantasies and explicitly presents naked bodies and sexual activities for the purpose of sexual arousal (Williams, 1989; McKee et al., 2020). Regarding media ethics and media effects, pornography has traditionally been viewed as highly problematic. Pornographic material has been accused of portraying sexuality in unhealthy, morally questionable and often sexist ways, thereby harming performers, audiences, and society at large. In the age of the Internet, pornography has become more diverse, accessible, and widespread than ever (Döring, 2009; Miller et al., 2020). Consequently, the depiction of sexuality in pornography is the focus of a growing number of content analyses of both mass media (e.g., erotic and pornographic novels and movies) and social media (e.g., erotic and pornographic stories, photos and videos shared via online platforms). Typically, pornography’s portrayals of sexuality are examined by measuring the prevalence and frequency of sexual practices or relational dynamics and related gender roles via quantitative content analysis (for research reviews see Carrotte et al., 2020; Miller & McBain, 2022). This entry focuses on the representation of relational context of sex as one of eight important dimensions of the portrayals of sexuality in pornography.