{
  "annotation_id": "9a442c3a-94a2-43f1-accf-a2ab2a8ec34a",
  "date_created": "2026-03-02T18:00:56.932000Z",
  "date_modified": "2026-03-02T18:00:56.932000Z",
  "file_hash": "3eeed29438f0b69115d549aece14d9b58d4c6629827b70f757d093024db9a9a7",
  "private": false,
  "record": {
    "abstract": "Although the inference of global community structure in networks has recently\nbecome a topic of great interest in the physics community, all such algorithms\nrequire that the graph be completely known. Here, we define both a measure of\nlocal community structure and an algorithm that infers the hierarchy of\ncommunities that enclose a given vertex by exploring the graph one vertex at a\ntime. This algorithm runs in time O(d*k^2) for general graphs when $d$ is the\nmean degree and k is the number of vertices to be explored. For graphs where\nexploring a new vertex is time-consuming, the running time is linear, O(k). We\nshow that on computer-generated graphs this technique compares favorably to\nalgorithms that require global knowledge. We also use this algorithm to extract\nmeaningful local clustering information in the large recommender network of an\nonline retailer and show the existence of mesoscopic structure.",
    "arxiv_id": "physics/0503036",
    "authors": [
      "Aaron Clauset"
    ],
    "categories": [
      "physics.data-an",
      "cond-mat.dis-nn",
      "physics.soc-ph"
    ],
    "doi": "10.1103/PhysRevE.72.026132",
    "journal_ref": "Phys. Rev. E 72, 026132 (2005)",
    "title": "Finding local community structure in networks",
    "url": "https://arxiv.org/abs/physics/0503036"
  },
  "schema_id": "dorsal/arxiv",
  "source": {
    "execution_id": "7827e259-e444-4ace-838f-2ae6aca8ff76",
    "id": "arXiv Dataset IDs",
    "type": "Model",
    "variant": "snapshot-2026-03-01",
    "version": "0.1.0"
  },
  "user_id": 1000002
}