Increasingly online reviews are relied upon to make choices about the purchases and services we use daily. Businesses, on the other hand, depend on online review sites to find new customers and understand people's perception of them. In order for an online review community to be effective to both users and businesses, it is important to understand what constitutes a high quality review as perceived by people, and how to maximize quality of reviews in the community. In this paper, we study Yelp to answer these questions. We analyze about 230,000 reviews and member interaction ("votes") with these reviews. We find that active and regular members are the highest contributors to good quality reviews and longer reviews have higher chances of being popular in the community. We find that reviews voted "useful" tend to be the early ones for a specific business. Our findings have implications on enabling high quality member contributions and community effectiveness. We discuss the implications to design of social systems with diverse feedback signals.

Online recommendation communities, like Yelp, are valuable information sources for people. Yet, we assert, review communities have their own dynamics behind the social interactions therein. In this work, we study the Yelp review votes of useful, funny, and/or cool to understand these social perceptions of the review. We examine the relationship between these social signals and the emotional valence of the review itself (text and rating). We aim to understand the community's perception of each of these signaling contributions. We construct a conditional inference tree of social signals from 230K Yelp reviews to study how social signals shape the deviance in review rating from the mean rating, an indicator of the overall business rating on Yelp. We find two effects of social signals. First, reviews voted as useful and funny are associated with lower user ratings and relatively negative tone in the review text. Second, reviews voted as cool tend to have a relatively positive tone and higher ratings. Our findings open a research direction for further understanding of perceptions of social signals and have implications for design of recommendation systems.

Online recommendation sites are valuable information sources that people contribute to, and often use to choose restaurants. However, little is known about the dynamics behind participation in these online communities and how the recommendations in these communities are formed. In this work, we take a first look at online restaurant recommendation communities to study what endogenous (i.e., related to entities being reviewed) and exogenous factors influence people's participation in the communities, and to what extent. We analyze an online community corpus of 840K restaurants and their 1.1M associated reviews from 2002 to 2011, spread across every U.S. state. We construct models for number of reviews and ratings by community members, based on several dimensions of endogenous and exogenous factors. We find that while endogenous factors such as restaurant attributes (e.g., meal, price, service) affect recommendations, surprisingly, exogenous factors such as demographics (e.g., neighborhood diversity, education) and weather (e.g., temperature, rain, snow, season) also exert a significant effect on reviews. We find that many of the effects in online communities can be explained using offline theories from experimental psychology. Our study is the first to look at exogenous factors and how it related to online online restaurant reviews. It has implications for designing online recommendation sites, and in general, social media and online communities.

We present an end-to-end measurement method for the detection of traffic shaping. Traffic shaping is typically implemented using token buckets, allowing a maximum burst of traffic to be serviced at the peak capacity of the link, while any remaining traffic is serviced at a lower shaping rate. The contribution of this paper is twofold. First, we develop an active end-to-end detection mechanism, referred to as ShaperProbe, that can infer whether a particular path is subject to traffic shaping, and in that case, estimate the shaper characteristics. Second, we analyze results from a large-scale deployment of ShaperProbe on M-Lab over the last 24 months, detecting traffic shaping in several major ISPs. Our deployment has received more than one million runs so far from 5,700 ISPs.

We propose an active probing method, called Differential Probing or DiffProbe, to detect whether an access ISP is deploying forwarding mechanisms such as priority scheduling, variations of WFQ, or WRED to discriminate against some of its customer flows. DiffProbe aims to detect if the ISP is doing one or both of delay discrimination and loss discrimination. The basic idea in DiffProbe is to compare the delays and packet losses experienced by two flows: an Application flow A and a Probing flow P. The paper describes the statistical methods that DiffProbe uses, a novel method for distinguishing between Strict Priority and WFQ-variant packet scheduling, simulation and emulation experiments, and a few real-world tests at major access ISPs.

We present an end-to-end active probing methodology that creates frequency-domain signals in IP network paths. The signals are generated by periodic packet trains that cause short-lived queueing delay spikes. Different probers can be multiplexed in the frequency-domain on the same path. Further, a signal that is introduced by a "prober" in one path can cause a crosstalk effect, inducing a signal of the same frequency into another path (the "sampler") as long as the two paths share one or more bottleneck queues. Applications of the proposed methodology include the detection of shared store-and-forward devices among two or more paths, the creation of covert channels, and the modulation of voice or video periodic packet streams in less noisy frequencies. In this paper we focus on the first application. Our goal is to detect shared bottleneck(s) between a "sampler" and one or more "prober" paths. We present a spectral probing methodology as well as the corresponding signal processing/detection process. The accuracy of the method has been evaluated with controlled and repeatable simulation experiments, and it has also been tested on some Internet paths.