Navicrawler
Nawicrawler is a web-based extension programme that registers link relations between websites. The relations are activated manually, that means that the link relations are only registers when you actually click on a hyperlink that links one site to another. These link relations can later on be transferred to Gephi to draw a cartographic visualisation of what websites links to each other and to what extent they co-link. This tells us about how websites, belonging to both institutional, governmental, non-governmental and business associated actors, follow a certain politics of association on the web.
EMBARKING POINT AND BORDERS IN OUR CRAWL
In our investigation, we have embarked from a Google search on "Ipanema Coffees", the coffee production company and plantation that has been the local centre of our investigation of the controversy on fair trade certification of plantation grown coffee. In that way we ourselves has localised a controversy that is rarely localised by other important actors of the controversy.
In drawing the borders in our research object, we have tried to include all relevant actors in the controversy. The causal question to this remark of ours would then be, how we made borders for our map, that means how we have chosen to excluded some actors. In regards, we have made some general selections of what actors to include and exclude. In example, we have included all of the national departments of FLO in our network despite that they don't all necessarily articulate a stand in the controversy of fair trade certification of planation grown coffee themselves, but only represents the stands of FLO due to their embeddedness in their standards. However, our inclusion of these actors serves an important point in regards to the visualisation of the controversy. Without the extensive web of nation departments, FLO wouldn't appear as the global actor it is in the network, and the visualisation would thus be misleading. On the other hand, we have chosen to exclude actors, like all of the fair trade certified retail coffee sellers linked to by national FLO actors and most of the retails linked to by Fair Trade USA. Some of the ones we have included, like Equal Exchange and Starbucks, however, play an important role in our network. Our thoughts behind including and excluding certain coffee retails is grounded in our previous research and reading about the fair trade movement and important actors herein.
GENERAL REMARKS ON MAPPING ON THE WEB VIA NAVICRAWLER
Generally, we have noticed that while American and European actors co-links to each other and news articles that verifies the arguments in the controversy of these actors, Brazilian websites doesn't use this kind of online politics of association to the same extend. Furthermore, there can be link relations that Nawicrawler doesn't recognise. We have identified that one of the reasons behind this can be that some websites are using shortlinks and that Nawicrawler isn't always capable of registering the redirection process between the shortlink and the actual site.
EMBARKING POINT AND BORDERS IN OUR CRAWL
In our investigation, we have embarked from a Google search on "Ipanema Coffees", the coffee production company and plantation that has been the local centre of our investigation of the controversy on fair trade certification of plantation grown coffee. In that way we ourselves has localised a controversy that is rarely localised by other important actors of the controversy.
In drawing the borders in our research object, we have tried to include all relevant actors in the controversy. The causal question to this remark of ours would then be, how we made borders for our map, that means how we have chosen to excluded some actors. In regards, we have made some general selections of what actors to include and exclude. In example, we have included all of the national departments of FLO in our network despite that they don't all necessarily articulate a stand in the controversy of fair trade certification of planation grown coffee themselves, but only represents the stands of FLO due to their embeddedness in their standards. However, our inclusion of these actors serves an important point in regards to the visualisation of the controversy. Without the extensive web of nation departments, FLO wouldn't appear as the global actor it is in the network, and the visualisation would thus be misleading. On the other hand, we have chosen to exclude actors, like all of the fair trade certified retail coffee sellers linked to by national FLO actors and most of the retails linked to by Fair Trade USA. Some of the ones we have included, like Equal Exchange and Starbucks, however, play an important role in our network. Our thoughts behind including and excluding certain coffee retails is grounded in our previous research and reading about the fair trade movement and important actors herein.
GENERAL REMARKS ON MAPPING ON THE WEB VIA NAVICRAWLER
Generally, we have noticed that while American and European actors co-links to each other and news articles that verifies the arguments in the controversy of these actors, Brazilian websites doesn't use this kind of online politics of association to the same extend. Furthermore, there can be link relations that Nawicrawler doesn't recognise. We have identified that one of the reasons behind this can be that some websites are using shortlinks and that Nawicrawler isn't always capable of registering the redirection process between the shortlink and the actual site.
Gephi
Gephi is an open-source programme enabling multi functional visualisations of files from digital methods programmes like Nawicrawler and ANTA. In the visualisation, home pages, documents or actors, are shown as nodes operating in a network connected by lines. Obviously, we have used this programme to visualise our Nawicrawler file of link relations between actors in our controversy. More subsantially, we have also used Gephi's Degree function to tell us what nodes are biggest; that means what nodes do most other nodes refer to and back. The appearance of large nodes thereby tell us something about what actors that articulate the controversy of fair trade coffee from plantations the most. The continuing visualisations of our Nawicrawl sessions has thus given us some navigation marks in what actors are central in the controversy.
However intangible and visually messy it might appear, we have chosen to draw one map over the controversy instead of separating it into several maps which might had made a more neat result. This decision is based upon two notions of ours: 1) in dealing with a global issue, making separate maps for separate geographical entities is not a sufficient way to visualise the dimensions of the controversy, 2) separating actors of i.e. production, standard-making and certification will neglect the fact that all of these kinds of actors are intertwined in the same controversy and by splitting them apart, we would create a misleading visualisation on the fact that we're dealing with a controversy going on in a multiplex network where all of these actors voices are important at the same time and in their relations to each other.
In our visualisation, we've categorised our nodes through nine different categories. While this might make it hard to distinguish the colours of different actors, it provides the viewer with a sufficient and not over-simplifying overview of what kind of categories involved in the controversy.
However intangible and visually messy it might appear, we have chosen to draw one map over the controversy instead of separating it into several maps which might had made a more neat result. This decision is based upon two notions of ours: 1) in dealing with a global issue, making separate maps for separate geographical entities is not a sufficient way to visualise the dimensions of the controversy, 2) separating actors of i.e. production, standard-making and certification will neglect the fact that all of these kinds of actors are intertwined in the same controversy and by splitting them apart, we would create a misleading visualisation on the fact that we're dealing with a controversy going on in a multiplex network where all of these actors voices are important at the same time and in their relations to each other.
In our visualisation, we've categorised our nodes through nine different categories. While this might make it hard to distinguish the colours of different actors, it provides the viewer with a sufficient and not over-simplifying overview of what kind of categories involved in the controversy.
Wordle
Wordle is a tool for making and visualising word clouds. You can encode the text you want to analyse via pasting it in a analyse field or through copying the URL address from the website which text material you want to analyse. Unlike Lippmanian Device, Wordle doesn't require configurations of keywords nor does it analyse full domains, and thus enables you to perform text analysis of smaller pieces of text without predetermining what words to look for. However, a major disadvantage in Wordle is that it's not able to tie together terms consisting of two or more terms (i.e. "small farmers" become "small" and "farmers").
ANTA
ANTA is a text analysis programme that encodes semantic relations between pdf. files and words or terms. In that way it can do the same sort of semi-automised semantic analysis as Google Scraper or Wordle, but it has some preferable features if you:
1) Don't want to pre-define your research objects in form key words (which you have to do in Google Scraper)
2) Want to restrict your analysis to a limited amount of text (whereas Google Scraper analysis all text on a whole web domain)
3) Want to use a more intelligent tool able to detect when more words belong to the same term (whereas Wordle is not able to register terms consisting of two or more words)
4) Want to register how many times a certain document mentions a specific term and how many times a specific term is mentioned in how many documents.
Furthermore ANTA has different visualisation possibilities than Wordle or Google Scraper. Thus, it does not make word clouds, but maps generated by Gephi. This enables you to get a visualisation of which of the analysed pdf. files using what words and, by using the In-Degree function in Gephi, get an overview of which of the words that are linked to, and thus used, the most.
We have used ANTA to analyse what words that are most commonly used in FLO and Fair Trade USA's respective standards regarding coffee production specifically.
1) Don't want to pre-define your research objects in form key words (which you have to do in Google Scraper)
2) Want to restrict your analysis to a limited amount of text (whereas Google Scraper analysis all text on a whole web domain)
3) Want to use a more intelligent tool able to detect when more words belong to the same term (whereas Wordle is not able to register terms consisting of two or more words)
4) Want to register how many times a certain document mentions a specific term and how many times a specific term is mentioned in how many documents.
Furthermore ANTA has different visualisation possibilities than Wordle or Google Scraper. Thus, it does not make word clouds, but maps generated by Gephi. This enables you to get a visualisation of which of the analysed pdf. files using what words and, by using the In-Degree function in Gephi, get an overview of which of the words that are linked to, and thus used, the most.
We have used ANTA to analyse what words that are most commonly used in FLO and Fair Trade USA's respective standards regarding coffee production specifically.