SEXI 2013

Proceedings now available:
Proceedings of the Workshop on Search and Exploration of X-Rated Information (SEXI 2013)

The proceedings include a summary of the workshop published in SIGIR Forum Spring 2013, overviews of the keynotes, and the papers presented at the workshop.

Maryanne L. Fisher: What We Know about the Sexual Side of Human Nature
and
Susanna Paasonen: Ubiquitous yet filtered: porn and the search

One of the major shifts in psychological research within the past few decades has been the application of evolutionary principles to human behaviour. By using evolutionary theory, scholars have learned far more about human nature than ever before. Specifically, this paradigm shift has changed how psychologists view human sexuality, and with it, how we understand human behaviour more generally. For example, we know much more about sex-specific mating strategies, or competition for mates. This knowledge means that we also know more about which topics most people are sexually interested in, and why. In terms of acquiring insights into behaviour, we have many tools at our disposal; we can ask people, perform eye-tracking studies, or use the argument that when we see an artifact created by humans, we are also viewing something about human nature. That is, just as the consistent types of information presented in tabloids reveals our evolved human nature, so do the frequency of specific search terms used to find pornography.

After briefly introducing “evolutionary psychology,” I will review various findings about the frequency and diversity of sexuality related search terms and discuss why they are so prevalent.

Maryanne L. Fisher is an Associate Professor in the Department of Psychology at Saint Mary’s University, where she is also a member of the interdisciplinary Women and Gender Studies Program. Her primary areas of research interest include women’s competition for men and understanding human universals in popular culture. She is the lead editor of “Evolution’s Empress: Darwinian Perspectives on the Nature of Women” (OUP, 2013), writing a book about women’s same sex competition, and is currently planning a book about variable in women’s sexuality.

On the one hand, it is commonly known that search terms for pornographic content top the uncensored listings of most popular search terms. On the other hand, such terms are constantly filtered out when such listing as published. Similarly, Google image searches are filtered by default as a “moderate” SafeSearch, and anyone who wants to access sexually explicit images needs to turn off the filter. SafeSearch filters out images that “contain pornography, explicit sexual content, profanity, and other types of hate content.” This phrasing equates sexually explicit images with hate, while the notion of safety associates them with risk and danger. This rhetorical move reflects the broader, paradoxical position of online porn as perpetually popular yet seen as problematic filth. This presentation questions the conflation of porn with notions of danger. Separating between moral and ethical argumentation in debates on pornography, I argue for the centrality of contextual understanding of the genre, its production, distribution and consumption.

Susanna Paasonen is professor of media studies at University of Turku, Finland. Specialized in internet research, cultural studies and studies of sexuality, she is most recently the author of Carnal Resonance: Affect and Online Pornography (MIT Press, 2011) and coeditor of Working with affect in Feminist Readings: Disturbing Differences (Routledge 2010) and Pornification: Sex and Sexuality in Media Culture (Berg 2007).

Adult content is pervasive on the Web, has been a driving factor in the adoption of the Internet medium, and is responsible for a significant fraction of traffic and revenues, yet rarely attracts attention in research.  The scientific community has spent considerable energy studying user-generated content and information access on the Web, to the exclusion of adult content.  This is understandable, as the topic is distasteful to some, and requires special legal and ethical considerations when asking employees, contractors and students to analyze and process the data.

Furthermore, the methods that work for other types of information access behavior are assumed to work for all types of content, including adult content.  We propose that this is an incorrect assumption.  In fact, even core concepts such as relevance and diversity, which are fundamental to any application involving information seeking and access, are defined differently for adult content.   Beyond effectively presenting adult content when a person is searching for it, it is not always clear whether a query refers to adult content, as many common terms are used euphemistically.  It is extremely important for a search engine to understand this before serving such content to a person who is not expecting it.

We propose that the research questions surrounding adult content access behaviors are unique, and we believe interesting and valuable research in this area can be done ethically.

We seek a greater understanding of the particular issues in accessing adult content on the Web. The focus of the workshop will be to put this area of research on the agenda, and explore the basic research questions that should be addressed in the field, the types of data needed for research, and the barriers to doing research this area.  Due to the lack of attention to this area of research there are many open questions.

CLASSIFICATION Even researchers and search applications not interested in adult content will have to deal with it in order to avoid it—presenting adult content to innocuous searchers is clearly a massive failure both for the individual searcher as well as for the reputation of the service.  What are automatic methods for identifying adult content, in particular adult user-generated content?  How can we identify adult content in video, images, and text?  What is the best way to identify adult query intent, and deal with ambiguous requests?  What are the appropriate ad placement strategies in adult content?

ACCESS Access to adult content seems to require a different approach than the ubiquitous navigation search—with searchers exhibiting an exploratory information seeking behavior, characterized by a diverse set of relevance criteria.  How should adult content be ranked?  How should search, exploration, and recommendation be balanced?  How does searching adult content relate to search on adult chat sites and social networks?  Is there a benefit to personalizing adult content?

EVALUATION  Given the distinct nature of adult content and the diverse relevance criteria, appropriate evaluation is crucial.   What is a relevant result, and what are suitable metrics for relevance?   Is adult content a recall-oriented, or precision-oriented task?  What is the right level of evaluation—individual requests or whole search sessions?   What is similarity and diversity in adult content?  How important is the avoidance of failure, relative to success?  Are searchers for adult content more tolerant of non-relevant results?

ETHICS  What are the ethical issues in working with adult content in an academic environment?  What are the ethical implications for the search industry, given that it partly facilitates the online adult industry?  How can adult material be made available so as to promote responsible behavior through the whole chain from production to consumption?  Is adult user-generated content more ethical than professionally produced media?

We limit our discussion to legal adult content.  Topics such as identifying online predators, child pornography, or human trafficking are out of the scope of this workshop.  Although these are important issues, they represent a separate set of research questions.

The outcome will be to define a set of research areas to elucidate the special issues surrounding the access of adult content.  We will discuss a set of best-practices for working with this data in an academic environment, and propose a research agenda for the near future.  The results of the workshop will have the form of a jointly authored report  to be published in SIGIR Forum.

Presentations in the workshop itself will not include examples of adult content, images or video.