Blog of Sara Jakša

Profiling from Blogs and from Social Media

On the first of April, I had presented my cognitive science master thesis topic at the class. The slides can be found here (in Slovenian). Very short summary would be, that I am researching the individual differences in sharing the opinion on social media.

After the presentation, I was talking to my classmates. One of them mentioned, that if this means, that I am more careful with what I post on the internet. My reply was, that I am posting my blog (the one that I am reading right here). And I don't mind, if anybody tries to analyse me based on this.

But on the other hand, I don't really post things on social media (I really need to delete the last remaining ones, that I have). And for anything that I can get in stores here in Ljubljana, I am using cash. So in a way, I am more careful in what data am I leaving, just in a different way.

I also studied business informatics, and this gave me a glimpse of what people can do with the data. Once the data is cleaned in the databases/tables, there is a lot of information, that can be gleaned from comparing the users. It is how the basic recommendation systems work. You find people or groups of people, that have the similar evaluation of the same works. And then check, what other works did these people also rated high and then recommend it to people, that they did not rate it yet.

Another example if the information, that can be gleaned from the liking behaviour. There was a good article, that showed how personality, gender and other attributes can be discovered through liking. People with different individual differences like different things, and this can be used to discover things about people.

But in order for this to happen, the data has to be in sort of a structured form. The blog is not exactly the structured form (unless it includes a lot of semantic web features). It is unstructured text, with maybe some videos and podcast thrown in. Maybe a picture or more as well. This means, that there is more work needed, in order to get these data in such form (the search engines still do it). So no everybody will do it with everybody.

The other reason I will borrow is from Janor Lanier book titled Ten Arguments for Deleting Your Social Media Accounts Right Now. And this is the BUMMER principle. But the BUMMER only makes sense, if the companies can get some money out of it. On my blog, I have no advertisements and I doubt I will ever have it, so what is the point of doing it?

Does this mean, that there is no way that people will abuse it? I mean, if I have a person trying to target me directly, I am sure they will go through all the writings that I wrote, trying to find something about me. But I am not afraid of that. I just don't want to be just another entity in the database.

Which is why I don't mind sharing the info through the Indie-web, which blogs are. And there is another plus form my side, and this is that the content is under my control. Nobody can delete the blogs but me. And even if the servers go down and the country blocks my webpage, I still have my backup. I don't have thin on any website, which I do not control. I had already lost some of my data, because the website stopped working. But here it depends on whatever I want to continue paying for the domain and hosting and nothing else.