I'm reading papers regarding microblog credibility perception of non-text cues.

  1. Tweeting is believing? Understanding microblog credibility perceptions by Morris et al., 2012

  2. Microblog Credibility Perceptions: Comparing the United States and China by Yang et al., 2013

However, Yang et al.[2013] refer to some cues, such as location, number of followers, is a reply as metadata. Metadata is defined as 'data about data', but such cues seem normal data to me.

Examples of cues mentioned:

  • author location
  • contains URL
  • number of followers
  • is a retweet

Now I'm confused what exactly does metadata include in microblogs, and what is the difference between metadata and normal data in microblogs.

For example: the number of followers - it's a non-text data. But is it a data or metadata?

  • 1
    $\begingroup$ I have no idea what Morris et al. [2012] is. It doesn't seem to be this one, for example. Also, presumably the paper has some examples on what constitutes metadata in their case. $\endgroup$ Commented Sep 22, 2016 at 18:15
  • $\begingroup$ They only refer to features as non-text data, but others mentioned the features in the study as 'metadata'. That's why I'm confused what is the difference between metadata and normal data. Please refer to the table on Page 5 includes all the cues. $\endgroup$
    – Mystery
    Commented Sep 22, 2016 at 18:29
  • 1
    $\begingroup$ To expand on D.W.s comment. The thinking behind comments is that they are ephemeral, in the sense that there's no expectation that they won't be deleted in the future. Consequently, if a comment is of the form "this new information belongs in the post", then it should be added to the post. By the way, welcome to the site! $\endgroup$ Commented Sep 22, 2016 at 19:50
  • $\begingroup$ How is this a computer science question? $\endgroup$
    – Raphael
    Commented Sep 22, 2016 at 20:05
  • $\begingroup$ The term "metada" is used in Yang et al.[2013], they replicated part of Morris's study and noted the cues as "metadata". $\endgroup$
    – Mystery
    Commented Sep 23, 2016 at 12:08

1 Answer 1


Metadata is an infamously ill-defined term. Roughly speaking, it's trying to get at a distinction along the following lines: data is the "content" you care about; metadata is other information that is secondary or exists primarily for technical reasons (e.g., addressing information, tags, technical data to facilitate storage or communication).

A classical analogy is postal mail. The contents of the letter is considered "data". The addresses on the envelope might be considered "metadata".

Of course, it's just bits, so at some level it's all data. But when we talk about the data/metadata distinction, we're usually trying to draw a distinction between the main content that users care about, vs secondary information used to enable the system to work.

Again, the term "metadata" is vague, ambiguous, and ill-defined. Don't expect a precise technical definition of the term. Instead, if you plan to use this term in communication and its precise meaning is important, then make sure to state clearly what you mean by metadata and what you do/don't consider to be metadata.

See also https://freedom-to-tinker.com/2008/10/20/what-your-mailman-knows-part-1-2/ and https://freedom-to-tinker.com/2008/10/23/abandoning-envelope-analogy-what-your-mailman-knows-part-2/ for discussion about the use of the term "metadata" in the privacy context.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.