Keeping Up With Reddit: Use IFTTT, Not Google Alerts

If you consider just mainstream press coverage, you might come away with the impression that Reddit is a chaotic place full of jerks and that the information it aggregates is of no use to the serious researcher.

And you’d be a tiny bit right. Reddit does have jerks, for the simple reason that a community of that size can’t exist without having jerks. In 2015, Reddit had over 8.5 million users provide over 725 million comments on over 73 million submissions. I imagine it would be impossible, statistically speaking, to aggregate that much human interaction and not have some untoward behavior.

Similarly I imagine it would be difficult to aggregate 73 million submissions on a site as wide-ranging as Reddit and not have something turn up that would be of interest or of use to you. Your field of interest would have to be very limited.

Knowing the size of Reddit and knowing its level of activity, I assumed that there was information in there that would be useful to ResearchBuzz. I just had to figure out a way to winkle it out.

After six months of various experiments, I’ve finally settled on a useful way to monitor Reddit without too much hassle. My biggest takeaway? Google Alerts is not the right tool for the job.

Information Trapping Strategies for Reddit

Reddit is divided into categories called subreddits. Any registered user can make a subreddit and there are literally thousands and thousands of them. Metareddit lets you search for various subreddits. The nice thing about Metareddit is that you can search for a keyword it’ll show you subreddits that are somehow relevant to that keyword, not just the ones that contain the keyword in the title. Thus a subreddit search for cows finds not only cows and happycowgifs, but also ranching and vegan (and SlothMemes and Bioshock. Sometimes understanding the connection takes a little digging.)

It may be that as you explore the subreddits you may find that all you need to do is join Reddit and monitor a couple of particular subreddits. If you want to monitor for information on genomics, for example, then genomics and ClinicalGenetics might be all you need. In that case, get a Reddit account, join those subreddits, and enjoy! (The topic “Best Practices for Hands-On Reddit” is a whole ‘nother article.)

Alas, that won’t work for me. I’m interested in online databases and digital archives and so forth, which cover thousands of subreddits. If I tried to monitor just one or two, I’d miss so much.

I started my Reddit-monitoring with Google Alerts.

My Experiments With Google Alerts

I currently have 137 Google Alerts. Three of them are Reddit-oriented, and were generated after I did a lot of experimenting. With many of them I found that I got too much, non-useful results, or just weirdness. Finally I settled on three:

intitle:”new database” site:reddit.com

intitle:”new tool” site:reddit.com

intitle:archive site:reddit.com

Are you might notice from the syntax, these are very crude searches. The keywords I want in the title, on the site Reddit.com. Adding “new” to the keywords database and tool where the only useful way I could find to narrow down the search results. You’d think “archive” as a lone keyword would be overwhelming, but it wasn’t. Publishing information on archives is not a big part of Reddit, as far as I’ve noticed.

The results I got in my e-mail for the Google Alerts didn’t tell me too much; I generally had to dig.

That’s not to say these Google Alerts weren’t useful at all. I found several things – mostly related to popular culture – that I’m sure I wouldn’t have found otherwise. But I got usable results so infrequently that I was sure I was missing other content.

Enter IFTTT

If you’ve been reading ResearchBuzz for a while, you know about IFTTT. The site allows you to automate certain tasks usings its different Channels. A channel can be a site (like Reddit) or a technology (like e-mail) or a platform (like Blogger) or even an appliance (Samsung has channels for a refrigerator, robot vacuum, and washer!)

IFTTT stands for “IF This, Then That”. You put the channels together in “recipes,” so that if something happens on one channel, something happens on another channel. As you might imagine, this lets you set up a lot of time-saving tasks. And best of all, IFTTT is free!

I didn’t even have Reddit on my mind when I went to set up some recipes in IFTTT — but browsing through the channels, I noticed there was one for Reddit. And then I noticed how powerful it was.

The Initial Setup

I’m going to assume you know the basics of IFTTT. If you don’t, MakeUseOf has a very thorough guide, which should teach you all you need to know about making recipes.

Start with the Reddit channel, at https://ifttt.com/reddit. You’ll need to connect that channel, and to do that you’ll need a Reddit account. Make sure you’ve got an IFTTT account and a Reddit account, and that your Reddit channel is connected. When it’s all set up it’ll look something like this:

Screenshot 2016-05-08 at 13.29.34 - Edited

Underneath this screen will be a list of popular IFTTT recipes using Reddit. That’s fun to browse, but the really good stuff is at the bottom of the screen: the triggers and actions.

Every IFTTT has a list of Triggers and Actions. A trigger is something that happens on the channel (the “If”.) The actions are things that happen because of something happening on another channel (the “then”.) The IFTTT channel has a lot of triggers, but only two actions.

The trigger I use is New Post From Search. As the channel explains, “This Trigger fires every time a new post submitted on reddit matches a search query you specify.” There is a limit to 20 items per search so you need to get specific.

Let’s make a recipe. I want a recipe to send me an e-mail whenever a Reddit search mentions online museum. We’ll start with the trigger New Post From Search. I’ve chosen the channel and the trigger, so I need to enter my search term:

Screenshot 2016-05-08 at 14.10.40 - Edited

Note that entering your search term is absolutely all you have to do, though you can refine your search with Reddit’s search operators (more about that in a moment.)

(Wondering why I didn’t put the phrase online museum in quotes? Oddly, Reddit does not appear to support phrase search.)

Now you’ll need to create a trigger. Because I’m using this recipe to monitor Reddit for new resources, I’m having the results emailed to me. But if you look at Reddit’s channels you’ll see you have other options, as well. You could add information to a Google Sheet. You could add the links found to Pocket (though I recommend you be sure your search is getting good results before you do that.) You could even automatically tweet something the monitor finds, though again you better be sure your results are fantastic (and that you’ve got the steely nerve required to auto-Tweet links found by automated Reddit searches.)

I find email works best for me, and IFTTT thoughtfully has a quick template of the results that I’ll get back from my Reddit searches. I already have the Email channel set up in IFTTT, so adding this action to my recipe only takes a moment.

PROTIP: If you don’t want to get an email every time the search finds a resource, IFTTT also offers an “EMail Digest” channel that will send you resource roundups either daily or weekly.

The last step is Reddit showing you a summary of your recipe. Click “Create Recipe” and you’re good to go!

Screenshot 2016-05-08 at 14.11.44 - Edited

Just looking at this recipe, you’ve probably noticed the obvious thing that sets it apart from Google Alerts: you’re searching Reddit directly, not relying on Google to index everything perfectly. That alone is going to increase the quality of the resources you get.

But there’s a second step here that will make your search results even better: using Reddit’s own search syntax.

Tweaking Your Reddit Searches

This was a very basic Reddit search: online museum . We can do better than that.

Reddit’s got a list of advanced search operators at https://www.reddit.com/wiki/search. Quickly skimming this list will give you at least a few ideas for making your search better. Let’s hit the highlights.

Reddit Syntax

title: – Just what it says on the tin. Limits your search terms to the title field of a search. If you have general topic interests this is the most important search syntax you can use to keep from being overwhelmed.

site: – If you’ve got a search that’s giving you quality problems and you know your keyword is good, site can be your best friend. It’ll restrict the link you find in your search. You don’t have to just restrict your search to a particular domain; instead, you can search for a top-level domain like edu, gov, or a country code like uk. Try the following searches in Reddit:

title:genetics
title:genetics site:edu
title:genetics site:gov
title:genetics site:uk

See how different those are?

nsfw: Whether a post is considered “safe for work” or not. nsfw:no means you do not want sites which are not considered safe for work. I always set this one to no.

STUPID PET TRICK: You can do a standalone search for nsfw:yes without any other keywords or syntax if you’re just interested in the more — um — adult side of Reddit.

self: Lets you specify whether a post is a text post, or whether it’s a submission with an URL. I want posts with resources, so I set this to self:no. On the other hand, if you were looking for information or anecdotes about something, you might set this to yes. Searching for title:spina title:bifida self:yes might give you insight to what people living with spina bifida or otherwise connected to it are struggling with or want to know more about, while title:spina title:bifida self:no gives you resources and news stories. I recommend you experiment with this before you use it.

When I use Reddit’s search syntax, my search recipe looks a little different. Instead of just online museum, it’s title:online title:museum nsfw:no self:no. Testing that search on Reddit gets me a reasonable number of useful results. I finish creating the recipe and bam, I’ve got a new information trap on Reddit.

I’ve published this recipe on IFTTT at https://ifttt.com/recipes/416887-search-reddit-for-new-resources-about-online-museums. You can make your own copy and edit it to reflect the topics and ideas that you’re interested in as you explore using Reddit as a resource for links and information.

A Couple of Hazards

In the month or so I’ve been using IFTTT and Reddit, I have found a lot of useful sites. I’m getting more e-mail, of course, but since the titles are so clear, I’m sure most of the time when I follow a link it’ll be useful. Two particular things, however, are an issue:

Cascades: I call them cascades, anyway. Sometimes you’ll get a useful link multiple times. The Panama Papers database release is a great example. I saw a few mentions of it, then as the release date of May 9th got closer I saw more, and then an absolute flood a couple days before before the database release. I just delete them as the subjects of the e-mail alerts make it clear they’re duplicates.

Old Resources: Sometimes someone will post a resource that’s very old thinking it’s new and cool. If you’re just gathering resources this won’t matter, but I try to keep ResearchBuzz focused on fresher links.

(What about spam? – Personally, I haven’t seen much spam on the Reddit resources I get. I think my keywords are keeping most of it at bay. If I did see more I would revamp my keywords or consider using the site: syntax to restrict my searches to at least the .edu and .gov top-level domains.)

Why Don’t You Just Use an RSS Feed?

Reddit, unlike many sites, offers an RSS feed of its search results (You can see the one for the online museum search here). So why don’t I just put an RSS feed together instead of messing around with IFTTT?

There are a few reasons:

  1. Getting the results the IFTTT way means you’ll have more flexibility on how you can use them – sent to Pocket, auto-tweeted if you’re nervy enough, etc.
  2. The duplicates are easier to deal when you get them individually in e-mail versus when you get them scattered all over an RSS feed – at least to me.
  3. My priorities. My e-mail alerts are a higher priority to me than my RSS feeds, because they’re targeted. When I get alerts sent to my e-mail, I know I’ll read them first. If I put them in an RSS feed it’s not clear when I’d get around to reading them.

This Just Works Better Than Google Alerts

When I first started using IFTTT as an information trap, I had no idea it would work this well. I wish I’d cottoned to it a long time ago! If you’ve been interested in Reddit but could not work out a good way to monitor it, try IFTTT. You’ll save a tremendous amount of time and I’d be shocked if you didn’t find useful resources.

Do you like ResearchBuzz? Does it help you out? Please consider supporting it on Patreon. Not interested in commitment? Perhaps you’d buy me an iced tea. I love your comments, I love your site suggestions, and I love you. Feel free to comment on the blog, or @ResearchBuzz on Twitter. Thanks!

MIT Announces MITx Online Learning Initiative

MIT announced this morning an online learning initiative called MITx, which it describes as “a portfolio of MIT courses through an online interactive learning platform.” MITx is being combined with an MIT-wide research initiative on online learning and teaching.

Online learning is nothing new to MIT; if you’re aware of online education at all you probably know about OpenCourseWare from MIT, which makes notes, lectures, etc from 2000 courses available for free at http://ocw.mit.edu/index.htm (along with updates via an RSS feed; love you MIT.) From the announcement it seems like the MITx initiative is different from OCW in that it offers courseware and community and credentials.

The fact that MITx will “feature interactivity, online laboratories and student-to-student communication,” (from the announcement) and “allow for the individual assessment of any student’s work and allow students who demonstrate their mastery of subjects to earn a certificate of completion awarded by MITx” (also from the announcement) makes it sound like this is the next step between the raw content of OpenCourseWare (lecture notes, video lectures, transcripts, etc) and the complete package of an on-site, MIT education. (In one way it’s a tiny step, but in another way the idea of credentials being available through free courses like this is a huge leap.)

Not that I’m going to be able to investigate any time soon; a FAQ on MITx indicates that this new initiative will not launch until Spring 2012, and that will be in an experimental, prototype form. The same FAQ notes that OpenCourseWare is not going away (“OCW will continue as before: It will make course materials from across the MIT curriculum available to the world for free. There will be no reduction in the level of what OCW offers.”) and that, while credentials will be available through the MITx platform (“Those who have the ability and motivation to demonstrate mastery of content can receive a credential for a modest fee”) an MIT degree assuredly will not (“MIT awards MIT degrees only to those admitted to MIT through a highly selective admissions process.”)

The Tech has a somewhat more extensive overview of MITx (and I guess MITx really is “a brand,” though it’s disconcerting to think of it that way) and an article from the New York Times.

Find Webinar Listings

Seems like every time you turn around you hear the word webinar. All it means is a seminar or presentation that’s held on the Web, but to me it also means, “I’m sure there are tons of cool presentations out there and many of them are free, but I don’t think I’ll be able to find them all.” I feel better now that I know about WebinarListings at http://www.webinarlistings.com/.

This site lists Webinars. (Sometimes they’re not hard to figure out.) Hit the calendar link and you’ll get a nice list of upcoming events for the next week or so. Actually it’s too lists; a list of featured webinars and a list of basic ones. Between the two lists there were thirteen webinars listed a recent day, from “The Secrets and Lies Behind Social Media Success” and “Introduction to Team-based Authoring” to “Strategic Planning as Organizational Development” and “Planning Your First Webinar” (in case you wanted to get meta.) The featured listings seem to have a lot more detail, but event pages for both lists have the essentials (who, what, when, and how much) as well as a bevy of ways to share the webinar, as well as list it on your own calendars. Most of the webinars I looked at were free, though I did find one that was $79.

If you’re only interested in specific topics you can also search for a webinar. I did a search for Facebook and found six webinars between now and the end of the month.

The site invites webinar hosts to list their webinars here; you have a choice between free listings and $60 featured listings. Check out this page for details on how to submit a webinar. And if you want to keep up with changes to the site, WebinarListings does have a blog.

Peer 2 Peer University Starts Round Two

Thanks to Creative Commons for the heads-up about Peer 2 Peer University, which has announced its second round of free and open online courses. Read this and sign up quick, because the registration deadline is February 28…

What the heck is P2PU? The tagline for the site is “Learning for everyone, by everyone, about almost anything,” which should give you a good overview. The site, which is run by volunteers, is trying to create a source for high-quality, low cost education.

Currently the site is in its second phase of courses, which will run from March 12 to April 23. You can get the course list at http://www.p2pu.org/course/list. Courses offered include “Solve Anything! Building Ideas through Design,” “Managing Election Campaigns,” “Intro to Concepts in Behavioral Economics and Decision Making,” and “Climate Resilient Cities”. You’ll have to register on the site before you can sign up for the courses.

I don’t know what’s on tap after this next round of courses — pilot phase three? — but you can follow Peer 2 Peer U’s blog at http://blogs.p2pu.org/.

Harvard Launches Beta-Test Site for Digital Scholarship

The Harvard Office for Scholarly Communication has launched its public beta site for its open access repository, DASH (Digital Access to Scholarship at Harvard). It’s available at http://dash.harvard.edu.

You can search the site by keyword or you can browse by community. There are two communities available at the moment — Faculty of Arts and Sciences, and Harvard Law School. Harvard Law School has about 64 items at this writing compared to over 1400 for the Arts and Sciences.

I did a search for politics and got 357 results. The results ranged from a 1991 article, The Politics of the Estranged Poor, to a pre-peer-reviewed article that actually won’t come out until 2010, International Adoption: The Human Rights Position. Search results are default-sorted by relevance but can also be sorted by title, submit date, and issue date.

Click on the title of a search result item for more detail, including abstract, a link to the published version, cite-able link, author information, and citation. Of course there’s a link to the full article as well; all the ones I saw were in PDF format.

I saw an RSS link, but when I looked at it, it seemed only to be a feed for getting information on the latest materials added to DASH. That’s fine now, but could easily get overwhelming as more communities are added to the site. If ever an archive screamed for keyword-based RSS feeds…..

There are something less than 1600 items in the repository now, though I found plenty to browse through with my keyword searching. If you want more, however, be sure to go back to the DASH front page to find links to more collections, like HMScholar and the Open Collections Program.