Social media treats all users the same: trusted friend or total stranger, with little or nothing in between. In reality, relationships fall everywhere along this spectrum, a topic social science has investigated for decades under the theme of tie strength. Our work bridges this gap between theory and practice. In this paper, we present a predictive model that maps social media data to tie strength. The model builds on a dataset of over 2,000 social media ties and performs quite well, distinguishing between strong and weak ties with over 85 % accuracy. We complement these quantitative findings with interviews that unpack the relationships we could not predict. The paper concludes by illustrating how modeling tie strength can improve social media design elements, including privacy controls, message routing, friend introductions and information prioritization. Author Keywords Social media, social networks, relationship modeling, ties,

In the last decade, blogs have exploded in number, popularity and scope. However, many commentators and researchers speculate that blogs isolate readers in echo chambers, cutting them off from dissenting opinions. Our empirical paper tests this hypothesis. Using a hand-coded sample of over 1,000 comments from 33 of the world’s top blogs, we find that agreement outnumbers disagreement in blog comments by more than 3 to 1. However, this ratio depends heavily on a blog’s genre, varying between 2 to 1 and 9 to 1. Using these hand-coded blog comments as input, we also show that natural language processing techniques can identify the linguistic markers of agreement. We conclude by applying our empirical and algorithmic findings to practical implications for blogs, and discuss the many questions raised by our work. 1.

People who review products on the web invest considerable time and energy in what they write. So why would someone write a review that restates earlier reviews? Our work looks to answer this question. In this paper, we present a mixedmethod study of deja reviewers, latecomers who echo what other people said. We analyze nearly 100,000 Amazon.com reviews for signs of repetition and find that roughly 10– 15 % of reviews substantially resemble previous ones. Using these algorithmically-identified reviews as centerpieces for discussion, we interviewed reviewers to understand their motives. An overwhelming number of reviews partially explains deja reviews, but deeper factors revolving around an individual’s status in the community are also at work. The paper concludes by introducing a new idea inspired by our findings: a self-aware community that nudges members toward community-wide goals. Author Keywords Online communities, ecommerce, product reviews, text

Popular products often attract an astonishing number of user reviews. Sifting through them by the thousands can be quite a headache, one that has inspired solutions like unique phrase identification (e.g., Yelp) and helpfulness ratings (e.g., Amazon). While clearly useful, these techniques only capitalize on a small subset of the reviews, and cannot answer questions like, “do these people care about the same things I care about? ” Talking Points is our solution to these problems. It employs natural language methods to summarize thousands of user reviews into a navigable, browserbased interface. In addition to describing our novel algorithm for feature extraction and sentiment classification, this paper presents the results of an exploratory user study of Talking Points. Our study suggests that users explore reviews far longer with Talking Points than with traditional methods. More surprisingly, in randomized sessions users seemed persuaded to choose those products augmented with