This map displays the most specific topic for each city (whichever river is considered) according to the results of the Search Engine Research Pages.
The topics are built automatically based on the contents of scraped pages, through the clustering of text segments into classes, which are displayed in tab topics tree.
An index of localness for each scraped page is calculated based on the url's domain and page's language
The following map shows, for each city, the proportion of web pages assessed as local.
Consider either roughly equally-sized text segments or complete web pages.
NB: in table txt_segment the words have been lemmatized: search for e.g. 'mine' rather than 'mining' or 'tree' rather than 'trees'.
Examples of search:
To consult the documentation associated to this appplease check this site.