Research – Categorization of Content Hosted on Freenet
Western societies have recently been highly attentive to a part of the internet known as the darknet during the past few years. While mainstream media has focused on the criminal aspects of these invisible parts of the internet, research, on the other hand, has paid more attention to the darknet technology, known as Tor. A recently published research paper focused on a different part of darknet technology, known as Freenet.
The research tried to identify the types of content hosted on Freenet, as well as their distribution. Throughout this article, we will review some of the interesting results of this study.
What is Freenet?
Freenet is an adaptive peer-to-peer network application that permits the publication, replication, and retrieval of data while preserving the anonymity of both publishers and visitors. Peer-to-peer means that each node across the network, which has downloaded the Freenet software, is responsible for a portion of the tasks on the network. Each machine that downloads the software becomes a part of the Freenet network and stores files or a proportion of the files (e.g. web pages, audio, video, and other documents) for everyone to use.
The Freenet network operates on the “small world” network principle. When a user installs Freenet, the software connects them to a group of peers. Each of these peers in turn have their own peers. Through these peer-to-peers, small world connections, all of Freenet’s content is accessible to all users. This contrasts to the ‘client-server’ model that characterizes most surface internet traffic, where data is stored on a server and a client (someone using a device) requests that data (usually by clicking on a hyperlink).
Freenet’s peer-to-peer model is relatively slower than the surface net’s client-server model, as requests that are sent must search the network for content on other machines that happen to be connected to the network at the time of the request. On Freenet, information can be uploaded to the network, copied onto other computers, and remain indefinitely, regardless of whether the original content creator is still managing the content.
Freenet uses the same internet protocols as those used by Clearnet, and the application can be understood as a “special-purpose network on top of the internet.” After downloading the application, users utilize standard web browser interfaces to access files within the Freenet network and use HTML and Java to render web pages.
Types of content on Freenet:
The study utilized grounded theory analysis to understand the type of content on Freenet. Grounded theory is a multistage inductive process in which data is collected and analyzed in a reciprocal fashion. This method is well suited for the study’s exploratory analysis. The table below shows the type and percentage of content found on Freenet via this study.
Table (1) Typology and distribution of content on Freenet
The most common type of Freesite (Freenet sites) is flogs which presented 29% of the studied sample. These are Freesites that contain a wide variety of content that appear to primarily be about self-expression. Flogs can also be sites for news, ideology, coding, and many other topics. What makes flogs their own category is their personalization. Content posted on a flog was possessed by the site owner as ‘my content’ or was so eclectic that it was more about self-expression than providing information on a singular topic or theme. Flogs are analogous to blogs in the sense that they possess the three main characteristics of blogs in their original manifestation: they are composed of chronological posts; these posts contain text, links, and data in each post; and there is some personalized commentary about these links.
The next most popular category is child pornography, also referred to as child abuse images, presenting 27% of the studied sample. The high prevalence of child pornography lends some support to the claims in mass media about Darknets being a criminal space. Freenet may be a space for a certain type of offender, as studies of online child pornography offenders suggest that they are high functioning, relatively educated adults who tend to restrict their transgressing to the consumption of online images.
Flogs and child pornography links are the most popular Freesites on Freenet. As illustrated by the disproportionate amount of traffic associated with a small amount of child pornography websites on Tor, link frequency may not accurately reflect data volume. It may be that other types of Freesites, those that are archives or informational in function, may store more if measured in raw data.
The third most popular category is represented by empty links, which are addresses that when requested produce no content or do not return a page. The paper made a distinction between empty and dead links, which is essential for understanding the Freenet space. A dead link, in everyday Clearnet parlance, would imply a request for a file from a website and that file not being available for the public (the opposite being a site going live by a web administrator making the website available on the internet). Because Tor also runs on a client-server model like the Clearnet, this understanding applies to it as well. The peer-to-peer design of Freenet, however, requires a slightly different understanding of what an unanswered request means. Links can be empty or even be considered old because there are few computers currently online housing the requisite data and compiling the data may take too long. In these cases, the requests are terminated. The data may or may not be available, depending upon the number of computers which are currently online.
The remaining categories are fundamentally archiving or storing different types of information. The sites placed in the category of ‘News and Information’, for instance, are different in content from sites categorized as ‘Programming’, but similar as they are both archiving content. Accessing these sites will bring one to a page with text accompanied by data or links. The information on these sites is more focused or thematic than flogs, and they lack a flog’s personalization. They are fundamentally archival in function.
Even though this study is one of the most detailed studies delving into the content of Freenet, the representation of analyzed data is questionable, especially as only one researcher categorized the sample of Freesites. This greatly raises questions regarding the validity of categorization of Freesites presented via this study. Future research is needed to verify the types of content and their distribution presented in the study.