Recommender systems have been evaluated in many, often incomparable, ways. In this article, we review the key decisions evaluating collaborative filtering recommender systems: user tasks being evaluated, types of analysis and datasets used, ways which prediction quality is measured, evaluation attributes other than quality, user-based system as a whole. addition to reviewing strategies used by prior researchers, present empirical results from various accuracy metrics on one content domain...

article Free Access Share on PHOAKS: a system for sharing recommendations Authors: Loren Terveen Human-Computer Interface Research, AT&T Labs--Research, Florham Park, NJ NJView Profile , Will Hill Brian Amento Computer Science Department, Virginia Tech. Tech.View David McDonald Information and University of California, Irvine IrvineView Josh Creter Pomona College, Claremont, Calif. Calif.View Authors Info & Claims Communications the ACMVolume 40Issue 3March 1997 pp...

Under-contribution is a problem for many online communities. Social psychology theories of social loafing and goal-setting can lead to mid-level design goals address this problem. We tested principles derived from these in four field experiments involving members an movie recommender community. In each the participated were given different explanations value their contributions. As predicted by theory, individuals contributed when they reminded uniqueness specific challenging goals. However,...

Wikipedia's brilliance and curse is that any user can edit of the encyclopedia entries. We introduce notion impact an edit, measured by number times edited version viewed. Using several datasets, including recent logs all article views, we show overwhelming majority viewed words were written frequent editors this increasing. Similarly, using same measure, probability a typical view being damaged small but increasing, present empirically grounded classes damage. Finally, make policy...

Eli Pariser coined the term 'filter bubble' to describe potential for online personalization effectively isolate people from a diversity of viewpoints or content. Online recommender systems - built on algorithms that attempt predict which items users will most enjoy consuming are one family technologies potentially suffers this effect. Because have become so prevalent, it is important investigate their impact in these terms. This paper examines longitudinal impacts collaborative...

Open content web sites depend on users to produce information of value. Wikipedia is the largest and most well-known such site. Previous work has shown that a small fraction editors --Wikipedians -- do Other offered conjectures about how Wikipedians differ from other change over time. We quantify test these conjectures. Our key findings include: Wikipedians' edits last longer; invoke community norms more often justify their edits; many dimensions activity, start intensely, tail off little,...

Social matching systems bring people together in both physical and online spaces. They have the potential to increase social interaction foster collaboration. However, lack a clear intellectual foundation: nature of design space, key research challenges, roster appropriate methods are all ill-defined. This article begins remedy situation. It clarifies scope by distinguishing them from other recommender related techniques. identifies set issues that characterize space shows how existing...

Article Free Access Share on The dynamics of mass interaction Authors: Steve Whittaker ATT Labs-Research, 180 Park Ave, Florham Park, NJ NJView Profile , Loren Terveen Will Hill Lynn Cherny Authors Info & Claims CSCW '98: Proceedings the 1998 ACM conference Computer supported cooperative workNovember 1998Pages 257–264https://doi.org/10.1145/289444.289500Published:01 November 1998Publication History 128citation1,847DownloadsMetricsTotal Citations128Total Downloads1,847Last 12 Months81Last 6...

Member-maintained communities ask their users to perform tasks the community needs. From Slashdot, IMDb, Wikipedia, groups with diverse interests create community-maintained artifacts of lasting value (CALV) that support group's main purpose and provide others. Said don't help members find work do, or do so without regard individual preferences, such as Slashdot assigning meta-moderation randomly. Yet social science theory suggests reducing cost increasing personal contribution would...

For many topics, the World Wide Web contains hundreds or thousands of relevant documents widely varying quality. Users face a daunting challenge in identifying small subset worthy their attention.

The discovery of a person's meaningful places involves obtaining the physical locations and their labels for that matter to his daily life routines. This problem is driven by requirements from emerging location-aware applications, which allow user pose queries obtain information in reference places, example, “home”, “work” or “Northwest Health Club”. It challenge map personally due lack understanding what constitutes real users' places. Previous work has explored algorithms discover personal...

Wikipedia has rapidly become an invaluable destination for millions of information-seeking users. However, media reports suggest important challenge: only a small fraction Wikipedia's legion volunteer editors are female. In the current work, we present scientific exploration gender imbalance in English population editors. We look at nature itself, its effects on quality encyclopedia, and several conflict-related factors that may be contributing to gap. Our findings confirm presence large gap...

Most commonly used approaches to developing automated or artificially intelligent algorithmic systems are Big Data-driven and machine learning-based. However, these can fail, for two notable reasons: (1) they may lack critical engagement with users other stakeholders; (2) rely largely on historical human judgments, which do not capture incorporate insights into how the world be improved in future. We propose describe a novel method design of such algorithms, we call Value Sensitive Algorithm...

Recent studies have found that people interpret emoji characters inconsistently, creating significant potential for miscommunication. However, this research examined in isolation, without consideration of any surrounding text. Prior work has hypothesized examining their natural textual contexts would substantially reduce To investigate hypothesis, we carried out a controlled study with 2,482 participants who interpreted both isolation and multiple contexts. After comparing the variability...

Over the past decade, social network sites have become ubiquitous places for people to maintain relationships, as well loci of intense research interest. Recently, a new site has exploded into prominence: Pinterest became fastest reach 10M users, growing 4000% in 2011 alone. While many articles appeared popular press, there been little scholarly work so far. In this paper, we use quantitative approach study three questions about site. What drives activity on Pinterest? role does gender play...

Mobile crowdsourcing markets (e.g., Gigwalk and TaskRabbit) offer crowdworkers tasks situated in the physical world checking street signs, running household errands). The geographic nature of these distinguishes from online raises new, fundamental questions. We carried out a controlled study Chicago metropolitan area aimed at addressing two key questions: (1) What factors influence whether crowdworker will be willing to do task? (2) how much compensation demand order Quantitative modeling...

Explanations are important for users to make decisions on whether take recommendations. However, algorithm generated explanations can be overly simplistic and unconvincing. We believe that humans overcome these limitations. Inspired by how people explain word-of-mouth recommendations, we designed a process, combining crowdsourcing computation, generates personalized natural language explanations. modeled key topical aspects of movies, asked crowdworkers write based quotes from online movie...

Emoji are commonly used in modern text communication. However, as graphics with nuanced details, emoji may be open to interpretation. also render differently on different viewing platforms (e.g., Apple’s iPhone vs. Google’s Nexus phone), potentially leading communication errors. We explore whether renderings or differences across give rise diverse interpretations of emoji. Through an online survey, we solicit people’s a sample the most popular characters, each rendered for multiple...