Bio

I am a PhD candidate in the Management Science & Engineering Department at Stanford University and am a member of the Social Algorithms Lab. My research interests include social network analysis, machine learning, and causal inference.

Education & Certifications

  • MA, Harvard University, Statistics (2015)
  • BS, Ohio University, Mathematics (2012)

Contact

  • Academic
    kaltenb@stanford.edu
    University - Student Department: Management Science and Engineering Position: Graduate

Additional Info

  • Mail Code: 8610

Links

All Publications

  • Monophily in social networks introduces similarity among friends-of-friends NATURE HUMAN BEHAVIOUR Altenburger, K. M., Ugander, J. 2018; 2 (4): 284–90

    View details for DOI 10.1038/s41562-018-0321-8

    View details for Web of Science ID 000435523500012

  • Monophily in social networks introduces similarity among friends-of-friends. Nature human behaviour Altenburger, K. M., Ugander, J. 2018; 2 (4): 284–90

    Abstract

    The observation that individuals tend to be friends with people who are similar to themselves, commonly known as homophily, is a prominent feature of social networks. While homophily describes a bias in attribute preferences for similar others, it gives limited attention to variability. Here, we observe that attribute preferences can exhibit variation beyond what can be explained by homophily. We call this excess variation monophily to describe the presence of individuals with extreme preferences for a particular attribute possibly unrelated to their own attribute. We observe that monophily can induce a similarity among friends-of-friends without requiring any similarity among friends. To simulate homophily and monophily in synthetic networks, we propose an overdispersed extension of the classical stochastic block model. We use this model to demonstrate how homophily-based methods for predicting attributes on social networks based on friends (that is, 'the company you keep') are fundamentally different from monophily-based methods based on friends-of-friends (that is, 'the company you're kept in'). We place particular focus on predicting gender, where homophily can be weak or non-existent in practice. These findings offer an alternative perspective on network structure and prediction, complicating the already difficult task of protecting privacy on social networks.

    View details for PubMedID 30936539

  • PuzzleCluster: A Novel Unsupervised Clustering Algorithm for Binning DNA Fragments in Metagenomics CURRENT BIOINFORMATICS Siegel, K., Altenburger, K., Hon, Y., Lin, J., Yu, C. 2015; 10 (2): 225-231

    View details for Web of Science ID 000354786800010