Anupriya Ankolekar

Friendlee: A Mobile Application for Your Social Life

We have designed and implemented Friendlee, a mobile social networking application for close relationships. Friendlee analyzes the user’s call and messaging activity to form an intimate network of the user’s closest social contacts while providing ambient awareness of the user’ social network in a compelling, yet non-intrusive manner.

Friendlee: a mobile application for your social life

We have designed and implemented Friendlee, a mobile social networking application for close relationships. Friendlee analyzes the users call and messaging activity to form an intimate network of the users closest social contacts while providing ambient awareness of the user social network in a compelling, yet non-intrusive manner.

Foxtrot: a soundtrack for where you are

In this paper, we present a mobile location-aware and crowd-sourced audio application, Foxtrot, that allows people to share the sounds and music they enjoy and associate with a particular location. Foxtrot plays an automatically created radio-like stream of geo-tagged music, ambient sounds and comments left by friends and other people. We discuss some of the design considerations for Foxtrot and our approach to selecting and scheduling audio content for playback. In addition, we present a pilot study of Foxtrot, which indicates that a location-aware music system might indeed provide an engaging mobile experience for users.

Play it by ear: a case for serendipitous discovery of places with musicons

Current location-based services (LBS) typically allow users to locate points of interest (POI) in their vicinity but can detract from the users emotional experience of exploring a new location. In this paper, we examine how cues in the form of popular music (musicons) can emotionally engage users and enhance their experience of discovering nearby POIs serendipitously in unfamiliar places. The primary contribution of this paper is a field study, in which we evaluate the performance and emotional engagement of different types of audio-based cues for directing users attention to specific POIs. Musicons and mixed-modality cues performed close to visual and speech cues, and significantly better than auditory icons, for POI identification while creating a much more pleasant and engaging user experience. We conclude that cues for POI discovery need not always be as explicit as the baseline visual cues. Indeed, the most challenging cues, auditory icons, led to a heightened sense of autonomy.

Kalpana - enabling client-side web personalization

A growing number of websites are recognizing the value of personalization based on a users context and social network. As more websites become personalized, the resulting experience for users can be rather fragmented. We aim to facilitate a seamless Web personalization experience across websites by enabling personalization to take place at the client and thus allowing personal information about people to reside locally with people. If websites are to script a personalization experience that draws on information held by the user, it is imperative that this information be easily comprehensible by heterogeneous websites. In this paper, we demonstrate how Semantic Web technologies can be used to realize a vision of client-side Web personalization. The contribution of this paper is an architecture that demonstrates the feasibility of our approach and a prototype implementation that establishes its viability.

Getting Users' Attention in Web Apps in Likable, Minimally Annoying Ways

Web applications often need to present the user new information in the context of their current activity. Designers rely on a range of UI elements and visual techniques to present the new content to users, such as pop-ups, message icons, and marquees. Web designers need to select which technique to use depending on the centrality of the information and how quickly they need a reaction. However, designers often rely on intuition and anecdotes rather than empirical evidence to drive their decision-making as to which presentation technique to use. This work represents an attempt to quantify these presentation style decisions. We present a large (n=1505) user study that compares 15 visual attention-grabbing techniques with respect to reaction time, noticeability, annoyance, likability, and recall. We suggest glowing shadows and message icons with badges, as well as more possibilities for future work.

MET: An Enterprise Market for Tasks

Crowdsourcing platforms have been rapidly harnessed by organizations for business uses, but enterprises continue to use traditional hierarchical forms of work allocation within their own boundaries. Using crowd work models within enterprises requires addressing enterprise-specific concerns such as how to maintain focus on employeesâx80x99 primary work while providing the right incentives to perform crowd work, how to promote wide employee participation while preserving management oversight and control. We present a novel crowd work system for the enterprise, the Market for Enterprise Tasks (MET), that addresses these concerns via two novel features: an incentive system tied to real-world dollars and multiple means of indirect control offered to the management, especially the ability to limit the size of tasks performed in the market. We have deployed MET in several groups within a large IT enterprise and report on initial experiences.

Evaluating Mobile Music Experiences: Radio On-the-Go

Music has become an accompaniment to everyday activities, such as shopping and navigating. Although people listen to music in a context-driven manner, music recommendation services typically ignore where a user is listening to the music. They also typically select music based on a single seed song, rather than ordering a user’s created playlists for the best user experience. The contributions of this paper are three-fold: (1) We present a survey of 15 DJs of college radio stations to identify their heuristics in creating playlists for radio shows. (2) We present an experimental study design to evaluate various scheduling (track ordering) strategies for mobile music consumption in situ, which is used to (3) conduct a field experiment that compares the user experience of three scheduling strategies (tempo, genre and location) against the gold standard of a playlist created by an experienced DJ (This work was completed when Anupriya Ankolekar and Thomas Sandholm were both researchers, and Louis Lei Yu was a postdoctoral research fellow at Hewlett Packard Labs. The majority of the experiments were conducted during the summer of 2011. The authors are listed here in alphabetical order).

Your attention, please! Determining saliency of competing audio stimuli in natural scenarios

Perceptual saliency is a precursor to bottom-up attention modeling. While visual saliency models are approaching maturity, auditory models remain in their infancy. This is mainly due to the lack of robust methods to gather basic data, and oversimplifications such as an assumption of monaural signals. Here we present the rationale and initial results of a newly designed experimental paradigm, testing for auditory saliency of natural sounds in a binaural listening scenario. Our main goal is to explore the idea that the saliency of a sound depends on its relation to background sounds by using more than one sound at a time, presented against different backgrounds. An analysis of the relevant, emerging acoustical correlates together with other descriptors is performed. A review of current auditory saliency models and the deficiencies of conventional testing approaches are provided. These motivate the development of our experimental test bed and more formalized stimulus selection criteria to support more versat...