Public | Automated Build

Last pushed: 2 years ago
Short Description
Short description is empty for this repo.
Full Description

btbServer

Data server for Beyond the Book

To start serving data, start the Flask server:

python btbFlaskServer.py

This will start the btbServer running on port 5000. The btbServer provides the following services:

Wiki contributions

http://localhost:5000/wikicontrib/<word>

Produce a list of Wikipedia contributions from various countries to the Wikipedia page of the given word. E.g:

http://localhost:5000/wikicontrib/London

Yields the contributions to the Wikipedia page for London. Contributions are given as an array in JSON format:

[
  {
    "expected": 0.383,
    "country": "US",
    "observed": 0.17480780603193377,
    "relative": -0.5112086342117579,
    "countryCode": "840"
  },
  {
    "expected": 0.132,
    "country": "UK",
    "observed": 0.5337670017740982,
    "relative": 0.7690804343009084,
    "countryCode": "826"
  },
...
  {
    "expected": 0.008,
    "country": "NL",
    "observed": 0.006859846244825547,
    "relative": -0.08169727675744143,
    "countryCode": "528"
  }
]

where each item represents the contributions from a country. Each item contains the following values:

  • country - 2 letter ISO code identifying the county
  • countryCode - 3 digit country code
  • expected - Expected percentage of contributions
  • observed - Observed percentage of contributions
  • relative - Calculated relative interest

Distances

http://localhost:5000/distances/

Provides the distance between books from the Sanders corpus using LIWC categories as the features to calculate the distance. An optional features parameter can be included in the request to select a subset of features to use. E.g:

http://localhost:5000/distances/?features[]=1&features[]=2&features[]=3

The result is a JSON structure containing a list of books, a square matrix of book-to-book distances, and a list of features used:

{
  "bookNames": [
    "2008_Jardin, Willem_Monografie van de mond",
    "1994_Haasse, Hella_Transit",
    ...
    "1981_Berk, Marjan_Nooit meer slank"
  ],
  "distances": [
    [
      0.0,
      0.018213146763971403,
      ...
      0.045538534384816205
    ],
    [
      0.018213146763971403,
      0.0,
      ...
      0.027707677598918585
    ],
    ...
    [
      0.045538534384816205,
      0.027707677598918585,
      ...
      0.0
    ]
  ],
  "featureNames": [
    "Category 2, I",
    "Category 3, We",
    "Category 4, Self"
  ]
}

Clusters

http://localhost:5000/clusters/

Produce a hierarchical tree of clusters from the Sanders corpus. Books are clustered based on their distances as described for the Distances.

As with the distances, an array for features can be supplied to select which features are used. Additionally, a maxdepth parameter controls the maximum depth of the hierarchical tree to be created. E.g:

http://localhost:5000/clusters/?maxdepth=3&features[]=1&features[]=2

will use features 1 and 2, with a maximum tree depth of 3 levels. Notice that nodes which could be further sub-divided are labeled: "Cluster containing N books", while terminal nodes are labeled with the title of the book.


{
  "children": [
    {
      "children": [
        {
          "children": [
            {
              "name": "2004_Noort, Saskia_De eetclub"
            },
            {
              "name": "Cluster containing 3 books"
            }
          ],
          "name": ""
        },
        {
          "children": [
            {
              "name": "Cluster containing 28 books"
            },
            {
              "name": "Cluster containing 13 books"
            }
          ],
          "name": ""
        }
      ],
      "name": ""
    },
    {
      "children": [
        {
          "children": [
            {
              "name": "Cluster containing 150 books"
            },
            {
              "name": "Cluster containing 167 books"
            }
          ],
          "name": ""
        },
        {
          "children": [
            {
              "name": "Cluster containing 89 books"
            },
            {
              "name": "Cluster containing 101 books"
            }
          ],
          "name": ""
        }
      ],
      "name": ""
    }
  ],
  "name": ""
}
Docker Pull Command
Owner
mkuzak
Source Repository

Comments (0)