The Dutch Blogosphere

Weltevrede, Esther, and Anne Helmond. 2012. “Where Do Bloggers Blog? Platform Transitions Within the Historical Dutch Blogosphere.” First Monday 17 (2-6). http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/viewArticle/3775.

The text below is a draft, please refer to paper above or the new project page: Dutch Blogosphere

Draft

Introduction

Lifeloggers and linkloggers

In 2003 Frank Schaap wrote an article on the Dutch blogosphere in which he noted the vibrant state of the Dutch blogosphere with the monthly e-zine About:Blank, its own Dutch Bloggies awards and two Dutch blog weblog systems (Pivot and ?). However, until then the Dutch blogosphere had been rather unknown to him due to the dominance of the English language on the internet. In his article Schaap makes several claims about the Dutch Blogosphere:

1. The Dutch blogosphere demarcates itself through linking policies in the blogroll: "Conversely, Dutch weblogs almost exclusively link to other Dutch weblogs in their blogroll, which only occasionally includes English language weblogs. In their entries, Dutch weblogs do link to English language web sites and copy or translate their news but the links of affiliation remain overwhelmingly with other Dutch weblogs. The Internet may be transnational but many communities remain bound by barriers of language."
2. The Dutch blogosphere looks at the English-language blogosphere but incorporates changes in a specific Dutch context, "showing the process of "glocalization" at work (Robertson 1995 in Schaap 2003)"
3. The Dutch A-list is small: "but there is a small set of some 50 to 75 high profile weblogs that get referenced and linked to a lot. This set of high profile weblogs consists mostly of weblogs that were among the first and second wave of weblogs, combined with some newer weblogs that have quickly gained popularity because of their particular format and content. However, few of the maybe two dozen "old school" weblogs, that were started in 2000 or even earlier, now remain, but a lot of the "second wave" weblogs, that were started in 2001, are still actively maintained. "
4. Schaap distinguishes Dutch lifeloggers from linkloggers:
5. Lifeloggers: 1) "Links to other weblogs, mentioning other webloggers, and commenting on what another weblogger wrote are means by which webloggers position themselves in the blogosphere. But lifelogs are about their authors." 2) "The lifelog may be located on the Web, but just as with many personal home pages it provides many links and details about the author's everday, offline life. Adding an "About Me" page, pictures, or maybe a webcam, reinforces those links." 3) "In fact, just like on the home pages earlier discussed, lifeloggers primarily present themselves through well-known categories like work, family, friends, musical preferences, place of residence, pets, and political opinions to name but a few." 4) "Because The Netherlands is such a small country, no one lives more than a 2.5 hour trip away, unless they happen to live abroad. I have as of yet only briefly explored the "back channel" and offline communication and interaction, but the latest meeting brought together some 50 webloggers."
6. Linkloggers: 1) "The situation is different on linklogs. Most of the high profile linklogs are not produced by one author but by a small group. The linklog authors of the high profile linklogs give out few or no personal details about themselves and use a pseudonym or, supposedly, their first name, but never their last name." 2) "Dutch linklogs always try to cause a stir, either by linking to information that already is the focus of some sort of row, or that will become the focus of a row because they are linking to it. This includes sensitive information about people, shocking and doctored pictures, pictures and links of hot babes but also hardcore porn, cool flash movies and games, and, quite popular as well, opendir links."3)"Although the authors provide a particular direction and give the linklog its unique slant, their personality is secondary to the linklog's genre of "programming," which is the basis for its hits and thus for its success."

Other claims:

1. Especially with link logs, there is a difference between popularity by hits and popularity by links.

Platformlogs

In its current state the blogosphere can no longer be seen as a separate sphere but instead should be studied in relation to social media platforms. The blogosphere is often defined as the interlinking of all blogs, or indexed and constructed by search engines. However, with the advent of social media platforms like Twitter and Facebook not only have the linking practices within the blogosphere changed, also the conversational aspect of blogging in comments have moved to social media platforms. This paper aims to contribute to the history of social media by analyzing the rise of social media platforms within media ecologies like the blogosphere by looking at changes in linking practices over time. It provides a case study of the Dutch blogosphere by looking at linking structure which anno 2010 is characterized by social media platforms.We have moved from a bloggers A-list to a platform A-list consisting of a top three of: Twitter, Flickr, YouTube

In ‘Links, Lives, Logs: Presentation in the Dutch Blogosphere’ Frank Schraap describes the state of the vibrant Dutch blogosphere in 2003. In his article he describes two types of blogs: the linklog and the lifelog. These two types of blogs have different linking practices: “Linkdumpers are understood to post primarily links to other websites, usually accompanied with a title and a brief comment from the author, while "lifeloggers" post primarily about their personal life and everyday experiences.” (Schaap 2003) Linklogs link to other blogs online while lifelogs link to their about page, personal pictures or offline context. In that sense their linking practice seems similar to that of the personal homepage that links to “offline contexts of hobbies, family, work, and place of residence.” (idem).

Anno 2010 we propose a new type of blog: the platformblog. This type of blog is characterized by embedding content from other social media platforms like Flickr, YouTube and Facebook and by referring to the author's presence on these platforms in sidebar widgets (see illustration 1).

Research Question

How have linking practices in the Dutch blogosphere evolved since 2003? When did the Dutch blogosphere move from linklogs to lifelogs to platformlogs? Are there other historical periods that can be distinguished analyzing the results? Can we identify different platforms belonging to different periods?

Method

Current blogosphere:

1. Create StartingPointsNLBlogosphere. The list is created from expert sources & two important Dutch log lists, the Loglijst and the Nedstat lijst. Sources:
   1. Loglijst 2003 - http://web.archive.org/web/*/http://weblog.lijst.nl/
   2. Meeuwsen, Frank. Bloghelden. Amsterdam: Bruna, 2010.
   3. Nedstat top 1000 weblogs' statistics - http://www.nedstatbasic.net/s?tab=4&link=3&id=837798&country=NL&category=3016
   4. Roze, Merel, and Chrétien Breukers. Schrijfwijzer voor het web. Amsterdam: 2009.
   5. Schaap, Frank, ‘Links, Lives, Logs: Presentation in the Dutch Blogosphere’, Into the Blogosphere: Rhetoric, Community, and Culture of Weblogs
2. Create hyperlink network using the Issuecrawler
3. Color code social media platforms

Archived blogosphere:

1. Create StartingPointsNLBlogosphere
2. Archive collection of the Dutch blogosphere using the Internet Archive.
3. The archive data set is periodized according to the following method: Periodization of the Internet Archive will happen through a k-means clustering algorithm. The algorithm allows to define the number of clusters (e.g. 4 periods corresponds to 4 clusters). The algorithm groups all time stamped urls into x clusters which are maximally coherent internally and maximally exclusive externally. Similar to results from mapping the A-list blogosphere.
4. A customized linkripper tailored to the Archive of the Archive is used to extract all outbound links from the archived archive by an archive-link-ripper. The tool only extracts links from the first page as pages on the second or third level in the archive might not be from the same archival date.
5. (Point of consideration: do we take all links into calculation or only use lists like blogrolls? Both are interesting but note that these produce very different maps. The all-links-option is random because we only take the posts that are on the front page of the blog; it is good because it not only shows a blogosphere of 'friends' but rather a blogosphere that can be characterized by its links to news or tech for example. Additionally, this analysis method more closely resembles the issue crawler analysis method. The blogroll only option is preferred because it is random. )
6. From the archived collection, fetch outlinks (1 iteration, 1 depth) and create a hyperlink network with Gephi. Repeat this step for each year.
7. Iteratively check the outcomes of the network whether they are blogs or not and subsequently use the blogs as starting points for next years' archive crawl.
8. As a check, use Frank Schaap's important historical blogs to see whether our method is correct, e.g. whether we found important historical blogs in the right year.
9. Use the expert lists to detect clusters (e.g. literary blogs, linkdumps).
10. Analysis: How do the hyperlink structures change over time? What type of clusters emerge? What can be read from the hyperlink structure? Are there distinct linkloggers and lifeloggers clusters? When do platforms and social media emerge in the network?

Archived blogosphere MIT7

text here

Analysis of the data & further research

Mapping the changing practice of blogging in the Netherlands

When did the Dutch blogosphere move from link blogs to weblogs to widgetized blog? Are there other historical periods that can be distinguished analysing the results? From the archived collection, fetch outlinks (1 iteration, 1 depth) and create a hyperlink network. Repeat this step for each year. How do the hyperlink structures change over time? What type of clusters emerge? When do platforms and social media emerge in the network?

Diagnosing the current condition of the 2003 Dutch Blogosphere

• Liveliness metrics: Are the 2003 blogs still online? --> fetch reponse code with http://marijn.digitalmethods.net/scrp/responder.php
• Freshness metrics: How fresh is the content on the blogs that are still online? --> fetch date stamp from RSS
• Hijacked metrics: Are the blogs still online parked, for sale or hacked? --> (need a rubust method for this)
• Popularity metrics: What is the current popularity of the blogs that are still online? --> fetch Pagerank, fetch traffic stats from alexa.com and trends.google.com
• Dated software metrics: How up to date is the software used in the blogs? --> (create method for this. List of historical software (e.g. platforms, browser, java etc)
• Creation data metrics: When where blogs first registered? --> fetch whois data
• Brokenness metrics: Do the links still work? use link validator

For a full description of the metrics, see https://wiki.digitalmethods.net/Dmi/KilledContent

Web Words

How did interface language change over time? Use the special collection downloaded from archive.org and create method to map this.

Linklogs versus Lifelogs

Lifeloggers have online meetings. Do they correlate with online Dutch blogopsheres?

http://wiki.bloghelden.nl/page/Weblogmeetings and http://www.rolandow.com/vp04/index.php

Findings

In a sample from the Dutch blogosphere we looked at the outlinks of the actors in the startlist and we found that Twitter, Flickr, Facebook, Hyves and other social media platforms appear as important actors within the network. In this sample of May 2010 Twitter is the dominant platform in the Dutch blogosphere receiving 34484 links from the crawled population.

In 2010 social media platforms receive the most links from the crawled population indicating their prominence on the web and in the blogosphere. Claim: We have moved from a bloggers A-list to a platform A-list consisting of a top three of: Twitter, Flickr, YouTube. The linking structure of the Dutch blogosphere anno 2010 is characterized by social media platforms.

In 2010 linking practices show the dominance of software platforms in the Dutch blogosphere. Twitter, Vimeo, YouTube and Flickr are central nodes in the network along with Technorati, Facebook, Digg, LinkedIn and Hyves.

Literature

Bruns, Axel. "Methodologies for mapping the political blogosphere: An exploration using the Issuecrawler research tool." Firstmonday. http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/rt/printerFriendly/1834/1718

• top335NLweblogs.ods: top335NLweblogs.ods

• bloghelden_platforms.pdf.jpg:
  

• bloghelden_dutchblogosphere.pdf.jpg:
  

• bloghelden_dutchblogosphere.pdf: bloghelden_dutchblogosphere.pdf

• bloghelden_platforms.pdf: bloghelden_platforms.pdf

Erik's notes

• Extracted linkurl from db dump, cleaned URL list to exclude non-existent and empty URLs. Retained 695 URLs.
• Looking for middle of year, as described here: http://web.archive.org/collections/web/advanced.html
  • but: http://web.archive.org/web/*/http://www.bonga.nu/speciaal/eigenhuisenlog/ vs http://web.archive.org/web/2001/http://www.bonga.nu/speciaal/eigenhuisenlog/ (gives oct instead of jun or aug)