Last year, curious to know a little more about the universe BD Spanish-speaking on the web, I had embarked on a mapping exercise of the sites of this community. By using free software, I had tried to concentrate on the Spanish sites of this theme.
With the help of Linkfluence, a research institute specialized in observing and analyzing trends and opinions on the social web, I try to put this on a larger scale and always with the same interest in the subject .
Scope of study
This time it is a more important corpus: 2,167 sites (931 sites in the first version of the study), still in Spanish but also in Spanish co-official languages. However, in this version, the preference of the location in Spain of the referenced sites has been abandoned for a greater representativeness.
The site crawl was carried out at the end of June 2011.
In the first version of the study, I used 4 open source tools: Navicrawler, Gephi, Flem and Flagfox to select, collect and map the sites listed.
In this enriched version, indexing of sources and mapping was done by Linkfluence (who is also part of the group of contributors who are behind the Navicrawler, Flem and Gephi tools). The indexing was carried out thanks to their robots or crawlers who examine the contents of each site and extract the hypertext links in a database then serving as reference for the cartography.
In addition to the automated data collection, the research institute has also made available its Flash card to manipulate all sites interactively.
To begin with, Linkfluence has crawled at 1 click of depth the web pages resulting from the corpus that I had constituted in the 1st study in order to extend it. In addition to the original 931, about 4,000 sites were discovered in this way. Based on indicators and connectivity scores developed by them, Linkfluence then conducted a pre-selection of potentially relevant sites for the study. After that, I took over to finalize the manual exploration, sorting and categorization of the new sites.
To circumscribe the corpus, I advanced step by step by proceeding by elimination.
After having discarded the so-called sites of the “upper layer of the Web” (Google, Facebook, etc.), visited and done a lot of cleaning among the “neighboring” sites (those outside thematic BD), I continued the sorting in Concentrating more precisely on non-Hispanic BD sites. Apart from some sites of publishing houses in French, widely referenced in particular by Spanish-speaking authors published there, all sites in other languages were excluded.
As for the qualification of the sources, I did the same work as for the first version of the study trying to classify the new sites visited in one of the 9 categories previously chosen: authors, publishers, associations & federations, Magazines & fanzines, comics, manga, information & chronicle, events, WEE network. In this process, the type of content has been privileged as much as possible (manga, comics, etc.), instead of the entity (authors, associations, etc.).
Details on maps
The nodes represent the websites and the links or edges joining them are the hypertext links that they exchange (the incoming links are in red, the outgoing ones in yellow). The size of a node depends on its degree of connectivity. Here, it depends on the number of links pointing to the node (inbound links). The more links there are, the bigger the knot. For details on the graphic conventions and method used to create the map, please consult its help section on the interface.
As with the previous study, the map below is not exhaustive either. The crawl performed was not done on the entire content of the web pages but at a click away from the home pages of the referenced sites. It obviously does not group all the Spanish-speaking BD sites and only refers to a fraction of the websites of the domain.
As for the proposed categorization of sites, it is personal and although the ranking has been carried out seriously, according to the rules set out in the above methodology, it is nonetheless debatable. The choice of a category for certain sites has sometimes proved to be a real puzzle as some are diverse!
Cartography and Analysis
On a very large scale, when one gradually tries to define the sphere to be studied and to define its boundaries, one discovers one’s neighborhood, the environment in which it flourishes. By doing this topological work, I could see