I’ve gotten quite a few requests from folks for a guide to livetweeting conferences, so here it is! Feel free to contact me if you have any questions or concerns. (I’m @rctatman on Twitter and Github.)
I use the TalkTweeter app (Android only) to automate thread my livetweets. You can get it here.
What is livetweeting and why do it?
In case you didn’t know, I regularly livetweet conferences I go to. You can see some examples of what this looks like on my Twitter, here.
Livetweeting can mean a lot of different things & everyone does it slightly differently, but for me livetweeting conferences is a way of taking notes on current research in a publicly-accessible format. My main goal is to accurately capture the main points of the authors’ work clearly enough that someone who wasn’t at the talk still has a good idea of what it was about. In addition to being a record of what talks I saw at a given conference (I probably go back and look at a talk I’ve livetweeted a couple times a month), I consider it service to the scientific community:
- It lets people attending parallel sessions (when two or more different sets of talks are delivered at the same time) know what’s going on in the other sessions.
- It helps authors disseminate their research.
- It allows researchers who can’t attend a conference I’m at, whether for financial or geographic reasons, to keep up with the current research and join in the conversation.
- It helps get research out to the public (albeit mainly the part of the public who follows nerds like me on Twitter).
- Having a semi-congruent text record of the main points of a talk improves accessibility for listeners who may have a hard time understanding the speaker due to, for example, hearing impairment or lack of familiarity with the author’s language variety. (Do note, however, that livetweeting is NOT A REPLACEMENT for signed languages interpretation.)
Especially recently, I’ve gotten a number of requests for help and tips on how to livetweet. It’s very much a learned skill, and it will take time to get better at, but below I’ve compiled some pointers that will help you livetweet conference talks.
The basic format I used for livetweeting is this:
- AuthorLastName: Title of talk #conferenceHashtag
- [as a reply to the first tweet] AuthorLastName: content of talk (my note: my comments) #conferenceHashtag
- Repeat 2 as needed
Each part of this formatting has a motivation.
- Threading tweets: Threading refers to a chain of tweets where each tweet is a reply to the one before it. This has several advantages. The first is that it keeps your followers’ feed tidy. Since Twitter automatically collapses the middle of long threads, you’re not taking up as much space as you would if you made each tweet individually. This also clearly maintains the chronological relationship of the tweets, so anyone can read the thread and follow the course of the talk.
- Putting the author’s last name in front of every tweet: since tweets can be retweeted individually, even if they’re in a thread, this allows others to retweet the parts they find relevant/interesting while still clearly crediting the author.
- Two authors: If the paper has two authors, I use the “LastName1 & LastName2”.
- More than two authors: If a paper has more than two authors, I use “LastName1 et al”.
- Starting with the title: This makes the content of the thread clear, and makes it easier for you to search later. (You can request and download your own complete Twitter history at any time.) Also, if the conference has a proceedings it makes it easier for someone to look up the specific paper.
- The conference hashtag: This organizes your tweets, helps others find the work if they’re interested in finding out more and lets other conference attendees find your tweets. If you’re quite positive that there isn’t one already established, you can start your own. The most popular format is the acronym for the conference and either the year it takes place (for example: #EACL2017) or the number of the conference (ex: #NWAV44). Hashtags are not case-sensitive.
You can use whatever format you like, though, as long as you’re consistent! To help people follow along, I have a quick summary of my tweet formatting conventions as my pinned tweet (one tweet that always appears as the top tweet of your Twitter profile).
I’m currently alpha-testing an Android app that automates some of my formatting, and I’ll be sure to share it once the developer’s ready to release it.
General guidelines
Here are some basic guidelines I try to keep in mind when tweeting.
- Each tweet should ideally be able to stand on its own, since retweets don’t have context & things that don’t require context are more likely to be retweeted.
- Try to imagine what you’ll want to remember about this talk in six months, and then tweet those things.
- Consider your audience (both in terms of your Twitter followers & who will be following the hashtag). Will they be able, for example, to figure out whether LDA is Linear Discriminant Analysis or Latent Dirichlet Allocation from context?
- Always tweet useful links the author put in their slides, especially code or datasets. Start typing these the moment they show up on the screen; people almost never leave links up long enough for you to take your time. Once you’ve got it, you can copy and paste it at your leisure.
- Make clear what are your comments and what are the author’s. If you have a lot of comments on a paper, you might consider retweeting/quoting the first tweet of the thread and start a new thread of just your comments/reactions.
- If you make a mistake (it’s ok, everyone does it), tweet a correction.
- Use a real keyboard. I can tweet from my phone, but it’s much harder.
- Have a paper program next to you so you can quickly double-check the spelling of names.
How and when to tweet
- Keep a second tab open to the hashtag for the conference (search for it on Twitter, then sort by “latest” rather than “top” and check in on it every so often to make sure you’re not duplicating someone else’s tweets.)
- You’ll usually have the most tweets at the beginning/end of a talk, for the context and the takeaway.
- Including pictures of slides (especially figures) can be faster than trying to type conclusions, but be mindful of avoiding including people’s faces, and be aware that unless you manually add an image description it won’t be accessible to folks using screen readers. (You can turn on the ability to add image descriptions in the accessibility portion of your Twitter options & I would recommend doing so. If nothing else, you’re helping to build a high-quality database of image captions ;)
- If someone asks a question on Twitter, I will (usually) relay it to the speaker and then tweet the reply.
There are some things that I make a policy of not including when I livetweet, and I would recommend you do the same.
- Don’t tweet pictures in which someone is identifiable without their consent. A good rule of thumb is not to tweet a photo with anyone’s face unless they tell you it’s ok. Some people don’t want their image on-line, and it’s far better to assume that that’s the case for everyone than, for example, unintentionally reveal someone’s exact location to an abuser.
- On a similar note, I sometimes replace the authors’ last name in a tweet with their Twitter handle (with a dot in front, .@twitterHandle, so that the whole name is shown in the tweet), but if and only if:
- I’m 100% sure it’s actually the author’s Twitter account. That’s not as easy as it sounds, especially for relatively common names!
- It’s easy to find by searching the author’s professional name on Twitter. (People with usernames that aren’t easy to link to their name usually have a reason for that.)
- It’s a professional Twitter account. You can usually tell if the top dozen tweets are all or mostly related to research.
- It’s relatively active. If they haven’t tweeted since 2010, linking their account to a talk they’ve given probably isn’t going to be helpful for anyone.
- Respect people’s boundaries, even if they’re unspoken. If you’re not sure, err on the side of not tagging someone.
- Don’t tweet preliminary or unfinished work.
- A good general rule is “Tweet unto others as you would have them tweet unto you”, and then err even a little more on the side of caution.
One of the unique challenges of tweeting as a medium is the strict character limit, and the format I use, with the addition of names (some of which are quite long) and hashtags takes a big bite out of the amount of space you have to work with. Add to that the time pressure of keeping up with the speaker, and you have a recipe for some pretty stressful editing. Here are some of the steps I take when I need to get a tweet under the character limit:
- Remove periods at the end of the tweet
- Make sure you don’t have any trailing spaces
- Drop function words like pronouns & articles (you can look at newspaper headlines for some examples of how this can be done)
- “the final model included…” → “final model included…”
- “then we fitted the model to…” → “model fit to…”-
- Replace “and” with “&” and “or” with “/” (where appropriate)
- “age and gender were both…” → “age & gender both”
- Instead of using commas for a list, split it over multiple lines. Newline = one character, space + comma = two characters
- Cut agentive & redundant language
- “the authors looked at the effect” → “the effect of X was”
- “feature selection for the model was done by…” → “feature selection =…”
- Cut your content over multiple tweets, but make it clear that there’s intended overlap (numbered list, ending the first tweet & starting the second with ellipses”
- “Three main findings = 1)…”, and then in the next tweet “2)…. 3)…”
- Finally, you can use abbreviations, but be careful about introducing ambiguity
- “including lg info improved…” Is “lg” language? Linguistic?
- Practice! I know it’s trite, but it’s true. You can try livetweeting talks from your field that have been posted on YouTube as a nice, low-stakes way to practice. Some channels that might be useful for this:
- Follow the hashtags of conferences you’re interested in, or look back at the ones from previous years. Notice what you find helpful & what you don’t.
Other resources
Gretchen McCulloch, well-known internet linguist and all-around cool person, has a great guide for getting started on Twitter & adivce on livetweeting here.
Joanna Bryson has a good discussion of some of the benefits (and possible drawbacks) of livetweeting in a blog post here.