Build a Social Network Map from a Text


This recipe extracts information about perceived social networks from a text populated with references to individuals.

  1. Take an electronic text from a source such as Project Gutenberg known to contain relationships between entities;
  2. Generate a list of words that indicate relationships using a tool such as WORDdetail Synonym and Antonym Finder; (Can the exercise do this with something like Romeo and Juliet…determine who’s related to who, which side one is on?)
  3. Prepare text by removing any added infrastructure;
  4. Use a List Words tool such as the Voyant Corpus Term Frequencies to get an idea of whether relation-indicating words occur in the text you are using;
  5. Use Concordance tool such as the Voyant Document KWICs on the text to with the relationship words;
  6. Take the results and use the collocation tool to assemble a list of relationships…..
  7. Could word tagging be useful here to indicating actions and objects of action?
  • Finding a Text

Possible sources for electronic texts are listed on the Electronic Texts Panel of TAPoR. When preparing text for analysis, you should be aware that academic infrastructure included in the text may obstruct reading the text for its original construction. It may be useful to remove notes and other materials added by subsequent authors from the original work. You can use tools such TAPoR Extract Text to remove added material.

  • WordNet Service

WordNet is one of many web services available which will provide word senses, synonyms, antonyms and other related words for terms that you enter. For more information see WordNet or the WORDdetail Synonym and Antonym Finder. In this case, to determine relationships involves distinguish between objects and people as well as between the parties relating to one another. To make this distinction…