How to seperate the "drive by" topics from the "deeper" discussions?

Community
21

I agree there’s plenty that could be automated to solve this and that this is an important aspect of what large social media companies do. But there are “manual” signals as well on social media, like who you choose who to friend/follow.

Within our organization, people opt in to some subset of a couple thousand Slack channels. That’s a pattern they are used to. (I’ll choose which topics/groups/categories I want to stay on top of).

I think there are important differences in Discourse, and one of the goals we have is to encourage cross-pollination, so I don’t really want to go down a road where people have to opt in to the things they want to see. And similarly, I don’t want to force people to mute things they don’t want to see.

I’d really like to keep the Latest page as the default “firehose” for those that want to see the full activity on the forum, while providing a simple and intuitive UX for people to build one custom stream of their favorite topics that they can focus on. I’d like this to work in a way that reduces noise, rather than boosting signal, which is why I’m not super keen on recommending that they just watch or track those categories.

I don’t object to exploring automated approaches, but I think there may be lower hanging fruit that is easier to reason about which could be tackled first.

This is definitely something that has also come up in our research so far as well.

6 Likes
22

hm. You’re both more experienced in managing communities than I am, but we’re talking here about a matter of degree, rather than a categorical distinction. I’m sure it’s true that relationships still matter in a forum, however my hypothesis would be that, the connection that ‘regulars’ have with a forum (and each other) is significantly more ‘topic-centric’ than it is on social media. I’m willing to be proved wrong, but I’d need some solid evidence.

Take meta for example. I could think there were cool and interesting people here (and I do :slight_smile: ). But if suddenly meta was re-purposed as a forum to talk about our favorite cars, then I probably wouldn’t come here that often.

There’s an analogy here with ‘work friends’. Work friends are people you know because you are pursuing the same goal. You naturally get to know them better in pursuit of that goal. You feel a sense of connection with them. However, when you change jobs your relationship with your work friends may last for a bit, but it often fizzles out. Maybe that’s just me.

I guess what I’m saying is that I definitely accept that relationships become more important over time, however the center of gravity is still going to be about specific subject matter. Conversely, on social media you care more about who it is that is saying or sharing something. Yes, what they are saying is still important of course, but who they are and what you think of them as an individual plays a greater role. You exist on social media as an ‘individual’ or a ‘personality’ (reflected in the ‘integrated’ experience @tophee mentions). You exist on a forum in a more targeted, focused way.

In some ways, allowing a user to manually curate their list is harder to do than ‘automatically’ making lists more user-centric. The ‘automated’ approach here would be to edit the algos in TopicQuery to preference topics similar to those a user has previously engaged on. You wouldn’t need to change much in the UI. The ‘automated’ approach also allows you to be more responsive to how people actually end up using your list. The problem with ‘manual’ approaches is that there is often a difference between what we think we’re interested in and what we’re actually interested in.

I think I will attempt the ‘automated’ approach to topic list curation based on proxies for user interest in a plugin at some point.

8 Likes
23

I think that’s a totally fair comment and I agree with you (although there are plenty of Facebook groups which fall outside the normal ‘social media’ formula).

What I’m trying (perhaps badly) to say is that I think that different types of forum audiences are motivated by different things – and by extension, different types of communities rely on different types of behavioural norms for their survival.

Agreed. The counter-argument is that if there was another community about the same topic which was filled with people that you have a lot more in common with, you’d be more likely to visit that one.

FWIW I’m really enjoying this topic and I’m not attempting to win any points – just picking your brains. I’m interested in your thoughts.

3 Likes
24

Since the vast majority of drive bys will be trust level 0 users, perhaps you could require trust level 1 to create a thread in any category other than tech support. When people get denied based on their trust level, you can remind them that anyone can post in the tech support category.

4 Likes
25

That definitely addresses the separation/getting lost in the noise issue but it will exacerbate the ‘we’re getting lots of crap and very few interesting discussions’ issue by raising the barrier to entry.

26

Addressing the good conversations getting lost in the noise should indirectly help bolster more discussion. Less frustrated users leads to less attrition which leaves you with more people to have interesting discussions.

I’m not sure the separation of wheat/chaff posts and the lack of interesting discussions are problems that can be fixed together with a single solution. Getting people to talk about interesting stuff on a forum is a much bigger problem than can be solved by tweaking category structure or user permissions.

3 Likes
27

Yes I agree. I think what cuts across that somewhat though, and what I see as the important connection between the UX of forums and social media, is the psychological aspect of product development; aka captology.

When Facebook is designing their news feed they don’t start from the assumption that their users are rational agents with definite preferences. They start from the assumption that their users have certain biological structures and responses that can be stimulated and even shaped over time. The UX decisions about the news feed are essentially based on neurobiology, not on higher level concepts of ‘interesting’ or ‘deep’. I’m sure that when they are making product decisions about their ‘groups’ features, they are taking similar things into account.

The ‘dark’ slant of this is thought that tech products are being designed to be addictive in the same way that drugs or slot machines are. The academic paper underlying that comparison is here if anyone is interested. A side note here is that that paper (and similar work) is often mis-reported. If you actually read the thing, they quite clearly say (a number of times):

technology “addictions” may not present the exact same brain etiology and possibly pathogenesis that drives substance and gambling addictions.

The more neutral reading of this direction is that of course neurobiology should play a role in UX and content curation. Taking into account the neurobiological responses of your users is simply acknowledging that they are human. The idea that there is some separation between a mental rational ‘self’ and our biology is a remnant of the dualist theme in Western thought going all the way back to Plato’s Forms. The question is not whether or not it is ethical to take a neurobiological approach to UX - all approaches have neurobiological consequences whether you realize it or not - the question is what are the ethics of the neurobiological approach you are taking.

This is partly what I meant when I said we need to think of the UX of social media as a product strategy, rather than a ‘practice intertwined with social and ideological narratives’. This is the context to why I think there has to be a role for things like automated user-specific curation in the discovery topic lists, rather than relying entirely either on user choice or user curation (i.e. moderation / community management). Automated user-specific curation is much better at incorporating learnings from advances in neurobiology / psychology, e.g. that there often isn’t a one-to-one relationship between our perceived and our enacted preferences; that we invariably seek the path of least resistance (we’re naturally lazy); that our ‘interests’ are not so much ‘chosen’, as they shaped over time by our environment; etc.

It doesn’t follow from this that Discourse should adopt practices like preferencing videos and images over text (like Facebook does). It flows from it’s topic-centricity that forum content, and engagement with that content, is different from social media.

It does follow from this that UX / product thinking about content discovery and curation on a forum needs to move more toward an explicit discussion of user psychology and away from approaches based the assumption that users are atomic rational agents with definite preferences, or that engagement on different types of communities is inherently different. People are behaviorally much more similar than they think. That behavioral similarity is how Facebook manages to ‘addict’ an academic at Yale as easily as a factory worker in Dhaka.

Finally, I would reiterate that I am not saying “Let’s think about how we can get users addicted to our forums”. I am saying that the real context to this discussion about content curation and UX is our neurobiological response to stimuli. ‘Addiction’ is a description of a state on a neurobiological spectrum. A spectrum which we all exist on, 100% of the time.

@HAWK none of that is directed against you per se. As you say, I think we’re broadly on the same page.

4 Likes
28

All good. I actually used to teach a captology/persuasive technology workshop (and how it relates to community platforms). I’m 100% on board with ^^

1 Like
29

And if try to make a plugin that adds a new page: feed
Which can be added to select pages (categories and latest + feed)

What to show on the feed?

1. bookmark
2. all the themes from categories, tags are not muted

The idea is that would make a Central page it is like the tape in social networks. This will allow to solve a lot of problems.

I just suggest if it make sense?

What I’m not signed, I won’t see. In social networks I do not unsubscribe from the content, and subscribe. Perhaps this approach is more General users. They see only what they want?

Although there is a “rule” of 3 clicks. In fact, each click is very important. For example. Now categories can be seen if you click on the tab. It’s one click. Transferring a list of categories in a bar we remove this with a click and do everything visually convenient. For example, I don’t like what I see, I suppose not satisfied with the topics from a specific category. I should go to this category and unsubscribe, that would not see the content on the main page of the forum. This extra action. Let originally that I like I will choose myself. As is done in all social networks. It certainly has a place when there are a huge number of people who post material on the forum.

2 Likes
30

Sorry but I disagree. Following your example, docs are docs because their calling is helping other. They are there because patients exist. Building a community where there are “ship classes” is not inclusive, something like “keep out the kids!”
I understand that docs should have a place where they can discuss “high level medical topics” but keeping the people out is not the right choice. Keep in mind that one of those “random people” (if you are kind to them) can become a doc himself, or even a chief.

1 Like
31

The avenue I would pursue is ranking content on a user by user basis by comparing content classifications - categories, tags and authorship (i.e. who made the content) - with user-interest proxies - likes, views, reading time, replies (and others) - of the individual user. Use that ranking to create a user-specific topic list.

Maybe you know how to already, but if you don’t, the way I would do it (and the way it’s done in @joebuhlig’s Hot Topics plugin) is:

  1. Add your list to the top menu options and the filter options is easy. Just add these lines to your plugin.rb e.g.

    Discourse.top_menu_items.push(:agenda)
Discourse.anonymous_top_menu_items.push(:agenda)
Discourse.filters.push(:agenda)
Discourse.anonymous_filters.push(:agenda)

  2. Add your list logic to TopicQuery. e.g.

    require_dependency 'topic_query'
class ::TopicQuery
  SORTABLE_MAPPING["agenda"] = "custom_fields.event_start"

  def list_agenda
    @options[:order] = "agenda"
    topics = create_list(:agenda, ascending: "true")
  end
end

I suggest you make both the content classifications and user interest proxies the subject of site settings so that different mixtures can be applied according to the initial hypothesis and experience of each forum.

4 Likes
32

Thanks, I tried to do:
https://github.com/Toxu-ru/discourse-feed-topics

The only thing that stops working on this page is sorting: Replies, Views, Activity
I think this approach can be useful.

1 Like
33

What we do in our community: we create topics a bit wider than the narrow scope of individual support questions, and then join the latter into the former.

Example “support” questions in the cheese home-making category:

  • My blue cheese turned out to be too hard, how do I soften it?
  • Blue cheese spoiled in my fridge. What are fridge requirements for blue cheese?
  • The ferment produced by XXX for blue cheese never worked for me, anyone tried it? What may I be doing wrong?

– all joined into a topic titled e.g. “Issues with blue cheese production”.

Then we have similar topics for soft cheese, hard cheese, cottage cheese and so on.

It has been working well so far. Just don’t make the combined topics’ scope too wide.

We also noticed that readers love such topics because they can read about many issues and read about experience of many people all in one thread – at least for cheese making it turned out to be useful to group discussions like these.

I’m not sure why it works. Maybe because when you cook something it is useful to read all the issues you can experience in one go in one place. Also, once an issue is solved (usually within <10 posts), the topic is over. If you leave it as is, the topic will get lost and never get any more activity unless it is specifically searched for).

3 Likes
34

Someone recently liked one of my posts in this topic, which prompted me to re-read it.

Re-reading a discussion from 7 years ago can be a strange and fun experience, both for the things you’ve forgotten you said, and also what the passage of time brings (e.g. I’m not sure that HAWK and Dave were working for Discourse in 2017, and are now, respectively co-CEO and Product Manager. This was also before I started Pavilion). One might say that the “discussion archeology” opportunities that forums afford are one of their strengths.

Relatedly, I’ve been meaning to look into the recent addition of the “hot” topic list (added in this commit), and re-reading the above gave some impetus. The notes in the code provide a useful equation for the hot algorithm:

(total likes - 1) / (age in hours + 2) ^ gravity

There’s some discussion of this addition in the topic announcing it

I want to preface my comments on the hot topic list with the observation that they’re essentially thought bubbles and that I agree with Lindsey’s overview of how the product utility of the Hot list (or any other topic list) should be determined:

That said, here are some things that come to mind out of the intersection of this old discussion and the new Hot topics list:

  • I wonder if “taxonomies” could be optionally used to weight the hot algorithm, i.e. a certain category or tag may get a lot of activity (e.g. likes), but you (the forum admin) consider that “shallower”, or less desirable in some way, than activity in another category or tag. In other words, adding in more site-specific discovery curation.

  • I wonder what a topic list contextualised to user variables (e.g. trust level, the prior topic views of that user) would look like / how it would behave. In other words, adding in more user-specific automated / algorithmic curation.

In other words:

4 Likes
35

I joined CDCK in July 2017 so I had just joined when I contributed to this topic the first time. I totally agree with your sentiment on how strange it can be reading your own old posts.

I like these general ideas and can think of lots of examples of situations where being able to curate hot to some degree would be useful.

3 Likes