Ajita John

What makes conversations interesting?: themes, participants and consequences of conversations in online social media

Rich media social networks promote not only creation and consumption of media, but also communication about the posted media item. What causes a conversation to be interesting, that prompts a user to participate in the discussion on a posted video? We conjecture that people participate in conversations when they find the conversation theme interesting, see comments by people whom they are familiar with, or observe an engaging dialogue between two or more people (absorbing back and forth exchange of comments). Importantly, a conversation that is interesting must be consequential - i.e. it must impact the social network itself. Our framework has three parts: characterizing themes, characterizing participants for determining interestingness and measures of consequences of a conversation deemed to be interesting. First, we detect conversational themes using a mixture model approach. Second, we determine interestingness of participants and interestingness of conversations based on a random walk model. Third, we measure the consequence of a conversation by measuring how interestingness affects the following three variables - participation in related themes, participant cohesiveness and theme diffusion. We have conducted extensive experiments using dataset from the popular video sharing site, YouTube. Our results show that our method of interestingness maximizes the mutual information, and is significantly better (twice as large) than three other baseline methods (number of comments, number of new participants and PageRank based assessment).

Social Synchrony: Predicting Mimicry of User Actions in Online Social Media

We propose a computational framework to predict synchronyof action in online social media. Synchrony is a temporalsocial network phenomenon in which a large number of usersare observed to mimic a certain action over a period of timewith sustained participation from early users.Understanding social synchrony can be helpful in identifyingsuitable time periods of viral marketing. Our method consistsof two parts – the learning framework and the evolutionframework. In the learning framework, we develop a DBNbased representation that includes an understanding of usercontext to predict the probability of user actions over a set oftime slices into the future. In the evolution framework, weevolve the social network and the user models over a set offuture time slices to predict social synchrony. Extensiveexperiments on a large dataset crawled from the popularsocial media site Digg (comprising ~7M diggs) show thatour model yields low error (15.2+4.3%) in predicting useractions during periods with and without synchrony.Comparison with baseline methods indicates that our methodshows significant improvement in predicting user actions.

Connecting content to community in social media via image content, user tags and user communication

In this paper we develop a recommendation framework to connect image content with communities in online social media. The problem is important because users are looking for useful feedback on their uploaded content, but finding the right community for feedback is challenging for the end user. Social media are characterized by both content and community. Hence, in our approach, we characterize images through three types of features: visual features, user generated text tags, and social interaction (user communication history in the form of comments). A recommendation framework based on learning a latent space representation of the groups is developed to recommend the most likely groups for a given image. The model was tested on a large corpus of Flickr images comprising 15,689 images. Our method outperforms the baseline method, with a mean precision 0.62 and mean recall 0.69. Importantly, we show that fusing image content, text tags with social interaction features outperforms the case of only using image content or tags.

Contextual Prediction of Communication Flow in Social Networks

We present a formal framework for media interpretation that leverages low-level information extraction to a higher level of abstraction in order to support semantics-based information retrieval for the Semantic Web. The overall goal of the framework is to provide high-level content descriptions of documents for maximizing precision and recall of semantics-based information retrieval.The paper develops a novel computational framework for predicting communication flow in social networks based on several contextual features. The problem is important because prediction of communication flow can impact timely sharing of specific information across a wide array of communities. We determine the intent to communicate and communication delay between users based on several contextual features in a social network corresponding to (a) neighborhood context, (b) topic context and (c) recipient context. The intent to communicate and communication delay are modeled as regression problems which are efficiently estimated using Support Vector Regression. We predict the intent and the delay, on an interval of time using past communication data. We have excellent prediction results on a real-world dataset from MySpace.com with an accuracy of 13-16%. We show that the intent to communicate is more significantly influenced by contextual factors compared to the delay.

Methodological reflections on a field study of a globally distributed software project

Context: We describe the methodology of a field study of a globally distributed software development project in a multinational corporation. The project spanned four sites in the US and one in India, and is a representative example of the complexities and intricacies of global corporate software development. Objective: Our goal is to provide the rationale behind the methodological choices and derive insights to inform the methodology of future studies of global software engineering teams. The paper also aims to provide an illustrative case of a typical geographically distributed corporate software project, through an in-depth description that emerged by applying the methods. Method: We reflect upon the reasons for choosing each of our methods, viz., non-participant observation, site visits, interviews, and an online questionnaire. We then discuss what we learned from the experience of applying the methods. Results: During and after the study, the discussions surrounding our methodological choices yielded important insights. The dynamics of software engineering practice and the geographical distribution of the project impacted factors such as access, costs, and cultural and linguistic diversity, and influenced the choice of methods. Our experience makes a case for methodological breadth and plurality as a means to a broad understanding of a global project. This understanding could then be linked to the specific research questions under consideration. Conclusion: The in-depth contextual description of the project that emerged from our methods highlights the utility of our methodological approach and provides an illustration of the complex nature of these projects. Our systematic reflection also yielded several methodological insights and provides important implications for future empirical studies of global corporate software development. Our experience can serve as a useful resource in methodological choices for research on globally distributed software engineering teams, or collaborative knowledge work in general.

XCHOKe: malicious source control for congestion avoidance at Internet gateways

The paper describes an algorithm to control flows from nonadaptive sources in the Internet in a simple and lightweight fashion. The algorithm, called XCHOKe, is an extension to CHOKe and RED. The paper presents experimental results from simulations that demonstrate that the algorithm outperforms current well-known algorithms for buffer management in providing better protection for TCP flows under varying degrees of attack from non-adaptive flows.

The future of online social interactions: what to expect in 2020

Blogs, wikis, tagging, podcasts, and social networking websites such as MySpace, Facebook, Flickr and YouTube have radically changed user interactions on the World Wide Web from a static, one-way, consumption model to a dynamic, multi-way, participation model. Broad user power and flexibility have changed how people engage in and experience their interconnections, interests, and collaborations. Online social interactions will evolve in the next decade to address the growing needs of its user community and make entries into many aspects of our lives. This evolution may very well be among the most exciting ones of our times where the individual and collective power of people to contribute and share content, experiences, ideas, expertise etc. may be even more enhanced than it is today. The enhancement may be shaped through a better understanding of user needs and behavior and it may be enabled through the seamless convergence of multi-modal technologies, new applications, new domains, data mining and better navigational and search capabilities. Some of these changes will also permeate into the workplace and change the way we work. This panel will discuss how online social interactions may evolve in the next decade and what impact it may have on diverse dimensions in our world.

Comparing privacy attitudes of knowledge workers in the U.S. and India

We compared privacy attitudes of knowledge workers from the U.S. and India who were involved in a collaborative software development project distributed across five sites of a multinational corporation. Prior studies on consumer privacy suggest that privacy concerns in India are lower than those in the U.S. While our work largely confirmed these findings, we found unexpectedly that knowledge workers in India expressed higher interpersonal privacy concerns compared with their U.S. colleagues. Our study points to a number of explanatory factors for the elevated privacy concerns in the Indian knowledge workplace: nature of interpersonal relationships, associations with privacy, competition among team members, management style and hierarchy, and differences in the physical characteristics of the workplace. Our findings highlight the challenges in satisfying privacy needs when individuals and teams collaborate with knowledge workers in India. An understanding of these issues is important for building and deploying systems for intercultural collaboration that can accommodate differences in privacy concerns.

Hermes: a platform for context-aware enterprise communication

This paper describes our vision for next-generation context-aware enterprise communications that achieves the integration of backend business processes with user communications. We envision a new class of enterprise applications in which communications between users in response to a variety of enterprise events will be driven by an automated process in which an appropriate group of users will be selected for communicating at an appropriate time on an appropriate media. To achieve this goal, the applications must exploit a variety of context information such as enterprise knowledge, user knowledge and application knowledge. This paper describes our platform Hermes that enables the creation and execution of such context-aware applications. It briefly presents experiences with Hermes in several demonstrations

PISA: automatic extraction of traffic signatures

Analysis of security attacks shows that an attack leaves its imprint or signature in the attack packets. Traffic from Distributed Denial of Service attacks and rapid worm spreads has the potential to yield signatures. While all signatures may not be indicative of attacks, it is useful to extract non-transient signatures that are carried by a sufficient number of flows/packets/bytes. The number of packets/bytes in the flows carrying the signature may be used for rate-limiting the flows, providing for timely and automated response to both known and unknown attacks. This paper proposes an efficient algorithm, PISA, which clusters flows based on similarity in packet information and extracts signatures from high-bandwidth clusters. Extensive experiments on two weeks of real attack data of 100 million packets yield about 1744 signatures. Additionally, PISA extracted the signature for the Blaster worm connection attempts in a mix of traffic from a trans-Pacific backbone link.