Social Computing Group

United States

Social Computing Group

United States
SEARCH FILTERS
Time filter
Source Type

Chua F.C.T.,Social Computing Group | Lim E.-P.,Singapore Management University | Huberman B.A.,Social Computing Group
HP Laboratories Technical Report | Year: 2014

Deep within the networks of distributed systems, one often finds anomalies that affect their efficiency and performance. These anomalies are difficult to detect because the distributed systems may not have sufficient sensors to monitor the flow of traffic within the interconnected nodes of the networks. Without early detection and making corrections, these anomalies may aggravate over time and could possibly cause disastrous outcomes in the system in the unforeseeable future. Using only coarse-grained information from the two end points of network flows, we propose a network transmission model and a localization algorithm, to detect the location of anomalies and rank them using a proposed metric within distributed systems. We evaluate our approach on passengers' records of an urbanized city's public transportation system and correlate our findings with passengers' postings on social media microblogs. Our experiments show that the metric derived using our localization algorithm gives a better ranking of anomalies as compared to standard deviation measures from statistical models. Our case studies also demonstrate that transportation events reported in social media microblogs matches the locations of our detect anomalies, suggesting that our algorithm performs well in locating the anomalies within distributed systems.


Abbassi Z.,Columbia University | Aperjis C.,Social Computing Group | Huberman B.A.,Social Computing Group
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2012

We have conducted three empirical studies of the effects of friend recommendations and general ratings on how online users make choices. We model and quantify how a user deciding between two choices trades off an additional rating star with an additional friend’s recommendation when selecting an item. We find that negative opinions from friends are more influential than positive opinions, and people exhibit “more random” behavior in their choices when the decision involves less cost and risk. Our results are quite general in the sense that people across different demographics trade off recommendations from friends and ratings from the general public in a similar fashion. © Springer-Verlag Berlin Heidelberg 2012.


Liu Z.,IBM | Jansen B.J.,Social Computing Group
Journal of the Association for Information Science and Technology | Year: 2016

Many people turn to their social networks to find information through the practice of question and answering. We believe it is necessary to use different answering strategies based on the type of questions to accommodate the different information needs. In this research, we propose the ASK taxonomy that categorizes questions posted on social networking sites into three types according to the nature of the questioner's inquiry of accuracy, social, or knowledge. To automatically decide which answering strategy to use, we develop a predictive model based on ASK question types using question features from the perspectives of lexical, topical, contextual, and syntactic as well as answer features. By applying the classifier on an annotated data set, we present a comprehensive analysis to compare questions in terms of their word usage, topical interests, temporal and spatial restrictions, syntactic structure, and response characteristics. Our research results show that the three types of questions exhibited different characteristics in the way they are asked. Our automatic classification algorithm achieves an 83% correct labeling result, showing the value of the ASK taxonomy for the design of social question and answering systems. © 2016 ASIS&T.

Loading Social Computing Group collaborators
Loading Social Computing Group collaborators