The newspaper approach to the discussion

Introduction

Finally, after analyzing the academic perspective, a new research field was explored: the journalistic one . This choice was due to two main different reasons. First of all because some newspaper and database of news were found in the second chapter, during the research on terms, and they were identified as category of actors. In second place because the research question is about knowledge and the field of journalism offers an interesting pony of view, as this media is recognized for the diffusion of information and so knowledge. According to the actors found in the second chapter, three newspaper were taken in consideration: New York Times, The Guardian and Google News.
Even if the comparison with Google News is a particular choice, as it’s not a real newspaper but just a news database taken from different sites, it still gives a different point of view as, first of all is an important actor of the controversy linked with the theme news, and second of all because it still collect news with a criteria and so shows a “point of view” about the debate.
The aim of this chapter is to create a chart that shows an tries to interpretate the approach of these three actors to the discussion. In fact, there is an analysis of the articles to understand which type of position they assume and to which category these articles belong. The gap of time analyzed is from 2010 to 2014 because in 2011 the book by Eli Pariser was published and it was interesting to see also the discussion even before the filter bubble theory was developed.

How to read the visualization

This visualization is an overall of the article found with the query “web personalization” on three online newspapers: New York Times, The Guardian and Google News. Each square represents an article and they are organize in groups by year and in order of relevance (as they were presented by the website).
The first element to identify the article is the color, that respects the code used for the last visualization: blue for the Pro personalization array and red for the Con personalization position. The grey shows the articles that weren’t relevant for the analysis and that we didn’t, for this reason, take in consideration: articles that were neutral, article that didn’t express an opinion on the topic, articles that didn’t talk at all about web personalization and pages that couldn’t be found because the database was too small (each one of this four groups are represented by a different symbol).
The last visual elements used for the chart are four geometrical shapes that identifies the article categories.

How it has been done

Using Kimono the research was set to find out different informations for each article: link, title, author, date and category (if it was indicated). These informations were found for all the first 20 results by relevance for each of the five years analyzed (from the of the 2010 to the 2014) for each newspaper, obtaining in this way a total of 250 articles per year (50 are missing from The Guardian).
With these informations three spreadsheet (one for each newspaper) was created and for each article was defined if it was Pro, Con or not relevant.

Findings

As well as in the rest of the analysis, it was possible to see the distributions of positions respect the article categories that are Technology, Social, Business and Media. The thickness of the circle represents the percentage of importance of that category on the total of that newspaper.
It’s possible to see that among the three group of articles there aren’t discrepancies considering the level of the percentages by category. The only relevant information that is revealed is that The Guardian has no business articles but has an higher percentage of Media articles than the two other newspapers. This particular thing, however, doesn’t give a strong clue as it probably depends by the database of the newspaper. Instead, is more useful to analyze how in these categories Pro and Cons opinions for each category are distributed. This is represented with a division of the two colors within the circle.
The really particular thing is that Google news business category is much more distributed in the Pro side, while the situation is more balanced for the New York Times. The next relevant thing displayed on the overall chart is the percentage of articles Pro and Con on the total. From this data it was possible to assume the "trend" of the three newspapers, discovering that the New York Times has a balanced number of Pro and Con articles (so it is a neutral situation), The Guardian has slightly more Con articles and Google News has more Pro personalization articles. Comparing the distribution of Pro and Con articles by year it’s possible to see that for the New York Times and The Guardian the trend is unregular during the five years analyzed. These fact can coincide with events, such as the Facebook change of personalization policy in 2010 that is reflected in number of Con articles in the same year for the New York Times.
Instead Google News seems to increase more the Pro personalization side. Maybe not for its own interests but because, as it is a database of news from different newspapers, is always update in terms of technology, technology that maybe are becoming each year even more efficient on the personalization aspects.

Metadata

Timestamp: 01/12/2014 - 19/19/2014

Data source: New York Times, Guardian, Google News

Related Protocol

Download data (4MB)