Let’s put labels on it

Hashtags! Tags! Labels! Call them what you will, they are here to stay. My favourite tag system is on Twitter as I found their tags were useful to find any information pertaining to a topic. In addition, you can subscribe to a hashtag so that if you are interested in an event with an official hashtag you can be sure you get all the tweets posted. This functionality seems to be missing on Instagram, where tagging seems to be a huge point of the service – I find this very strange. If you are interested in a hashtag, you have to search for it all the time!?

All the different services web-wide have different tagging methods and systems and databases. I don’t think we will be able to find one unified, correct way for tagging for Facebook, a business database, your photos on your computer and Instagram. The way one uses the tags on these various place change, there is a sub-culture. I looked into Flickr’s guidelines for tagging. It was lacking, with no clear guidelines. I think it might be a moot point anyway – people will not read user manuals for something they feel that they already understand, so unless there are admins continuously correcting the info, the users will tag what they feel is suitable. #newwyearseve, #newyerseve and #newyearseve will never be unified as long as people can make their own. I don’t think we will have a perfect system on any of the platforms.

Instagram is heavy on tagging and the amount and inventiveness of tags seems to have no end. When you read the tags, they are like sentences with their own story in addition to the text below the photo! I find it quite tiresome, personally.

Facebook public posts are spammed in the comments section with people tagging their friends, where I would simply use Share – send as message in Messenger. I see no value in many strangers seeing me telling them the name of my friend, and we are now both present in a public post. I try to avoid that. Most people might not be aware that the whole world is able to see public posts at any time, and all of their friends will see the post in their feed. However, if I do tell them – the answer is that it’s not a problem, they don’t have anything to hide. That’s not the point. People need to realise that the internet never forgets (it will never fully disappear) and you and your identity is worth gold. In other words, I am not too fond of Facebooks tagging system, but from a marketing perspective it’s a brilliant way to spread your message far and wide through unsuspecting sharers.

Flickr has a medium-sized tagging amount from what I have seen. I have searched for many CC licenced photographs. The tags that the photographers and guests have set helped me to find photos I found useful. Beware of “noisy tags” though: everyone who looks at a photograph will find different meaning in it. In the guidelines, I found nothing in terms of whether the tags should be of what you can see in the photo or what you feel about the photo or what it represents. I think such guidelines would be difficult to enforce and it wouldn’t be suitable – who is to say that a picture of two friends hugging should only have #hug and #people, but not #friendship or #love? Where do we draw the line? Can we apply #sunny, and most likely the automated Flickr-bot would apply #outdoors, as it did indoors for one of my own photos.

Noisy tagging, defined as inaccurate or false, is therefore highly subjective, I find. When I needed to find a photo of a smartwatch I searched for Apple, watch, smartwatch and wearables. I got slightly different results each time, maybe if I found a better photo of a Pebble smartwatch I could have used that instead. The tags sometimes lead you to other, maybe better, results than you originally thought you would find. Similar to Google when they suggest “did you mean …”, the tags help me find other results I didn’t know I was looking for. “Noisy” tags have also helped machine learning understand what a picture might be showing, not just a car but #happy #family in a #new #car. (Tang et al., 2009) For a good use of this, consider that Facebook is now using artificial intelligence to “read” aloud and interpret pictures for blind people! (Weiss, 2016)

Misspelled tags or weird word combinations I can easily see as noisy tagging, however I think that those tags will drown in the volume of other more correct tags. On social tagging sites, you often have word-clouds or “trending” lists and those are made by how often the word is tagged. As most people will not tag a word wrong over and over again I believe the noisy tags will be put into the background on their own. Another reason for not having pre-made tags to choose from is that user-created tags help create new words and concepts that might not already be in a dictionary. This is where tagging is powerful. (Matthews et al., 2010, pp.447) For example, my previously mentioned Twitter-use where a tag might be made and everybody accepts the tag as the official for that topic. #blacklivesmatter was such a tag, recently. (Anderson and Hitlin, 2016)

Summarized, I would say: keep the tagging to a level that is common for the platform and medium. Keep the tags useful and in context but allow some interpretation as well as factual tags, if suitable. Ignore the number of tags as long as the content of the tags themselves is useful.

 

 

Bibliography

Anderson, M. and Hitlin, P. (2016) ‘Social Media Conversations About Race’, (August), Pew research center. pp.1-35.

Matthews, B., Author, Jones, C., Author, Puzon, B., Author, et al. (2010) ‘An evaluation of enhancing social tagging with a knowledge organization system’, ASLIB Proceedings, pp.447.

Tang, J., Yan, S., Hong, R., Qi, G. and Chua, T. (2009) ‘Inferring semantic concepts from community-contributed images and noisy tags’, Anonymous Proceedings of the 17th ACM international conference on Multimedia, ACM. pp.223-32.

Weiss, T. R. (2016) ‘Facebook Photos on iOS Now Accessible to Blind Users’, eWeek, pp.1-.

Image: Copyright Petit_louis, https://flic.kr/p/jupZb9,  imaged slightly cropped, published under 2.0 Generic licence (CC BY 2.0).