2009-04-29
|
Integrating external (e.g. blog) discussions
|
David BourgetUniversity of Western Ontario
|
Hi Richard. I've just had a look around to see how well option 1 could work. It didn't see many posts that we could automatically connect with papers. Most posts which discuss a paper don't make that sufficiently clear in the title for a machine to get it, and I think that would leave too much work for authors to do. People don't even categorize their own papers, so imagine. Anyway, I was thinking about another system that would largely serve the same function. People could register their blogs for monitoring (subject to acceptance). They would map their tags to PhilPapers categories, and their posts would automatically be imported as "external posts" to the relevant forums. External posts would show a snippet of the post with a link to the blog to see the whole post. I'm not sure if we should allow replies to external posts--this would risk splitting discussion.
For the power user, this system might not be better than subscribing manually to one's favourite blogs. However, I think a majority of potential readers just don't have the skill or organization to do this. Also, one problem I find with using feed readers to monitor blogs is that you typically can't subscribe selectively based on topics. If you monitor many blogs, the stream of posts becomes overwhelming. PhilPapers would do that selection for readers. I hope there would be a strong incentive for people to register their blogs. This could potentially increase their readership significantly--we've got nearly 1000 people subscribed to the phil of mind forum now, for example.
A part of this system could be your second option. People could create external posts manually for all forums including article forums.
|
2009-04-29
|
Integrating external (e.g. blog) discussions
|
|
Interesting. I'd assumed you wouldn't want to import every philosophical post (from appropriate blogs) into these forums, at risk of overcrowding. But perhaps this could be done in a manageable way. In particular, it would seem less overwhelming if, rather than spawning a new thread for each newly detected blog post, each forum had a single thread dedicated to external links, and "external posts" relevant to the forum are all imported into this one thread. (I'm not sure how this would work in terms of the categorization hierarchy: would area forums inherit the links thread of each leaf, plus have a generic links thread of its own [in case someone tags a post with nothing more specific than 'mind']? Or would you allow only leaf-level tags? Would it matter if a post were imported into multiple threads? Could such redundancy be avoided?)
I agree that we should want to avoid splitting discussion.
P.S. Is it possible for people to unsubscribe from particular threads, whilst remaining subscribed to (the rest of) the forum? This would seem desirable, if you end up implementing anything along these lines, to allow PhilPapers users to "opt out" of external blog updates if they wish, without interfering with their other notifications.
|
2009-05-01
|
Integrating external (e.g. blog) discussions
|
David BourgetUniversity of Western Ontario
|
Well, I don't have a very good sense of how many posts per day we could expect if we covered all "professional quality" blogs. I'd say about 5 on average but I'm not a big blogger. As a pro blogger, what's your assessment?
I was thinking about area-level tags, because we don't have more specific forums (aside for paper forums). Putting all external posts for a given area in one thread would raise the problem you mention about aggregated forums, though we could always leave those threads in area-level forums only. The all-in-one-thread system would also require major changes in how we deal with threads vs posts. The idea is that one finds what interests one by identifying interesting threads. Threads provide a kind of "bird eye's view" of forum contents. Our various browsing and monitoring facilities, current and future, are geared toward that. If threads become internally heterogeneous the principle will break down. For that reason I think a better solution to overcrowding would be to be more selective in our inclusion of blogs. Blogs which produce lots of philosophy posts, in particular, would have to be highly relevant and of high quality to make it. (I think your blog would make it!) We would want to be selective anyway--even on the all-in-on-thread system.
Re de-subscribing from a particular thread, this is certainly doable. I will consider it. There's a bit of a technical challenge in implementing this efficiently, but not a big challenge. We were also thinking about taking off old threads from email notices. That would create an incentive for people to start new threads when a topic begins to fork. The proper use of this new system would be to create a subscription to a thread if one is interested to follow it. We might add a "subscribe to this thread" link in forum notices to make that easier. A more radical approach is to not send notices about follow up posts at all. I'm inclined to do that too, though I'm worried that there are many cases where a thread becomes more interesting after a few posts. Those threads wouldn't get the attention they deserve.
|