Research on Social, Mobile and Grid Computing

Emerging social, mobile and grid computing: Ubiquitous computing has changed business as individuals are using new generations of tools to connect. Now every voice has the potential to be heard. Social media tools (e.g. blogs, wikis, forums, electronic social networks, micro-blogs, tagging, reviews) are the current mechanisms amplifying individual voices and have exploded in popularity in recent years. Individuals can influence many others almost immediately.

While inherently personal, these tools are profoundly affecting the economic environment as firms struggle to adapt to the associated risks and opportunities. From the risk perspective, negative impressions (e.g. product defects, bad reviews) are no longer confined to a limited circle of a consumer’s direct contacts; instead, social media outlets distribute content rapidly and broadly, irrespective of truth. Firms are deeply concerned with the speed with which investments in brands and products can be undermined. Yet, social media also provides opportunities for firms to engage consumers (e.g. soliciting suggestions, vetting ideas), to identify problems, and to mediate consumer-to-consumer communication. Further, not only does the overwhelming volume of information exacerbate both risk and opportunity, cognitive processes are fundamentally shifting as newer social media relies less on reflective (e.g. a traditional restaurant review) influences and more on reactive (e.g. an immediate microblog from a mobile phone while sitting in the restaurant) inputs.

I focus on exactly this new domain--- combining computational tools with normative guidance for managing the business impact of social media. I combine multiple methods (qualitative, computer-based experiments, large-scale data analysis, optimization) across disciplines (marketing, computer science, information systems, operations research) to understand causal mechanisms.

Some current and recent projects are:

    Social media and customer dialog:
    (with John Gallaugher) Social media has created both new opportunities for customer interaction, but also new risks as firms struggle to manage the new media. We use a detailed case examination of Starbucks to provide a framework for managing new social media. A report on this research appeared in MIS Quarterly Executive, 2010, volume 9, number 4, pages 197-212. Download
    Quality of knowledge generation:
    (with Jerry Kane) Social computing tools have lead to online communities loosely organized to create unprecedented information artifacts like Wikipedia. However, what leads to quality production? We examine this question using empirical analysis of the complete history of 2+ million revisions to 15+ thousand articles on Wikipedia. [Contact me for a working paper.]
    Knowledge retention:
    (with Jerry Kane) Having created quality information artifacts, how to online communities retain high levels of quality? We examine this question using the detailed revision history of 2,065 featured articles in Wikipedia. A report on this research is at MIS Quarterly. Download
    Understanding interest in user generated content:
    (with Jerry Kane and Nick Lurie) While users can generate online content, why does some content create more interest than others? This question is particularly important for advertising supported sites. We create a panel of 2,026,992 revisions of 14,088 Wikipedia articles by 40,479 unique contributors and analyze the number of viewers the articles attract. A report on this research is at Marketing Science. Download
    Sequential grid computing:
    (with Ishwar Murthy, Saby Mitra and Sri Narasimhan). Grid computing is traditionally applied in the context of parallelizable tasks; however, not all tasks can be made parallel. We use stochastic optimization models to demonstrate grid benefits even from sequential jobs. A report on this research was published in the INFORMS Journal on Computing, vol. 23, no. 2, pp. 174-188, 2011. Download
    Mobile computing and social media:
    (with Nick Lurie) While user generated content from traditional computing tools has been largely reflective, mobile devices change the nature of content by encouraging reactionary responses. We empirically examine this shift using user generated restaurant reviews. A report on this research is at Marketing Science. Download
    Generalized assignment problem:
    (with Ishwar Murthy) The generalized assignment problem is a core problem in optimization. We introduce a novel reformulation which extends the size of assignment problems which can be solved. A report on this research is at Discrete Applied Mathmatics. Download