Library clips

Just had a thought if you could create an OPML out of a del.icio.us tag bundle.

Instead of sending a bunch of URL’s (tags) to someone, you could use an OPML outliner, to send someone a tag outline of a portion of your del.icious account.

Problem with del.icio.us is that the export feature is for the whole account, you can’t export a selection of tags, or export a tag bundle.

What you can do is view a HTML version of your account, or just a tag, or an intersection of tags, but this just shows the last 15 entries, and not an archive.
NOTE: To do this choose export, this defaults to your main account, simply add a tag to the end of the URL.

But even so a HTML version is not as good as the sharing, interchange, and extensible features of an OPML file.

This example below assumes that the tag bundle heading is a tag itself, like in Raw Sugar, where the first level heading contains all the items of the second level headings…see here.

- Blogs
     Weblogs for universities
     Weblogs in the classroom
     k-logs: share knowledge
     Syndicate your blog
     Keep user statistics with Feedburner
     Full-text or summary feeds
     RSS Demographics researched
     Usability of blogs
     Statistics on who reads blogs
     A Conversation on Blog Research
     Search incoming links
     A look at a few blog search engines
     Promote your blog
     RSS vs. email readership
- Blog_edu
     Weblogs for universities
     Weblogs in the classroom
+ Blog_rss
+Blog_km
- Blog_rsch
     RSS Demographics researched
     Usability of blogs
     Statistics on who reads blogs
- Blog_search
     Search incoming links
     A look at a few blog search engines
+ Blog_mkting

Note again that the tag Blogs (at top) has aggregated content from all the other tags, as it is the tag bundle heading.

So this is a portion of a del.icio.us account, under the tag bundle Blogs…this is what it could look like as an OPML outline.

I haven’t used an OPML outliner before, but I assume it would be possible to import the RSS of each del.icio.us tag into an OPML file, and arrange the tags accordingly.

This OPML file would have a URL that you could view at OPML search, or do with it whatever you please…it would be great to be able to create an OPML file within del.icio.us.

I haven’t got time to try this but I thought I’d put it out there…didn’t want to let it slip away.

Bloglines seems to be the only RSS engine which doesn’t generate a feed for citation searches (incoming links)…which is a pity as it probably currently does the best job.

A way around this may be to do a citation search on Bloglines, then monitor that search using a URL tracer like WatchThatPage or TrackEngine.

Firstly create a Bloglines Email Subscription, then join one of these monitoring services with that Bloglines generated email (I’m not sure if they will accept this email address), then do your search at Bloglines and hit your monitor service bookmarklet.

Hopefully any new content will be sent to your Bloglines RSS reader!

Pluck has a persistent search feature, I wonder if this would work, has anybody tried this?

[ADDED 12/05/06: Ponyfish : scrape a Bloglines citation feed]

For my job I’ve been tracking news using the different News Engines…lucky for me they all supply search RSS feeds.

News Engines I use:

Google News
Yahoo! News
Topix

…others MSN News, All Headlines News

…I also follow some of the RSS engines as well.

In a previous post I mentioned the difficulty in choosing which source to use or just use them all (and of course Factiva is another ball game).

A problem for me was duplication both within a feed and amongst all the feeds

…I did searches such as: Rio Tinto Australia, but also did other searches such as: Rio Tinto (and selecting Australia in the country news source field in the advanced section)…just to be exhaustive.

So for every news service I had 2 searches, times that by the number of news services (Google News, Yahoo! News, and Topix), and that’s 6 feeds for just one search topic.

If I had 5 search topics, that’s 30 feeds to look through every morning…and again the annoying thing was the duplication in each feed, and across all feeds.

Well the first great help was when Google News added RSS feeds (I was always using watered-down hacks), now every post in my RSS reader includes the collapsed related stories that Google does so well.

…and on top of that you could also get a feed for the whole of your Google Customised News Page, so this is kind of a spliced feed of all your customised sections (so that eliminated 6 feeds to 1).

Now I have a simple idea I haven’t tried yet that will hopefully reduce the duplication, therefore reducing the overload, and sparing me some time. Whack all my feeds into a re-mixer like Feed Digest and turn on the duplication feature
(Filter items with duplicate Titles, or filter items with duplicate URL’s).

This way I will only have to see the same story only once and I’m getting all my content from just one feed
…as for related stories (which aren’t technically duplication), at least from Google News they are being collapsed.

This is also handy for ego searching (checking your incoming links at all the RSS engines)
…instead of subscribing to feeds for incoming links from several RSS engines, just collaspse then into one, and set the duplicates…time saving and efficient…Talkdigger kind of does this.

TechCrunch have a great post on searchfox, an RSS Reader that learns from your reading behaviour…helps manage RSS overload.

From TechCrunch quoting searchfox:

“Existing RSS readers only show information chronologically, which quickly leads to information overload. Our goal is to that you see what’s interesting to you on the first page, rather than on the 20th page.”

From the searchfox blog:

“SearchFox uses machine learning technology to automatically rank and personalize incoming feeds to reflect each reader’s unique interests. Our RSS reader learns by watching what individuals and the entire community find interesting, taking into account various inputs such as the source, author, and topic of an RSS entry. Existing RSS readers only show information chronologically, which quickly leads to information overload. SearchFox makes RSS easier and immediately useful, so everyone can be part of the new web experience.”

See the demo.

Others:

Personal Bee
Findory
Chameleon for Bloglines
UltraGleeper

Current tag searching (blogs and social bookmarks)

tagbert is a tag engine similar to Tag Central, only it also includes content from blog categories.
The results are grouped in boxes from the services they are coming from, so it is a meta-tag engine but the results aren’t mixed in the main body.

Similar but more robust is Technorati Tags, this only lists content from blog categories in the main body, and content from bookmark services are listed on the sidebar…IceRocket Tags only has blogosphere content which is listed in the main body.

Tagosphere

The tagosphere is made up of:

- blog categories
- tags from social bookmark services

Needs:

- Blogosphere category engine (we already have Technorati Tags, and IceRocket Tags)

- Meta-Social Bookmark engine (Technorati could have a seperate service that aggregates tags from several social bookmark systems, and show the results in the main body, as it does for the blogosphere)

- Tagosphere engine (combination of both)

Next we need full-text search within a meta-social bookmark engine, this is what the dearly departed Gataga was trying to facilitate…Zniff does this for the Spurl social bookmarking service…but we need a meta-search service.

This is already a reality with the blogoshere…if we want to search the full-text within Technorati Tags (blogosphere categories), well then this is like searching from the standard Technorati screen…searching the blogosphere part of the web (and some news…can’t remember if they separate blog content from the main engine like Feedster does with blogs).

The human indexed-web:

Blogosphere full-text search (Technorati)
Blogosphere tags search (Technorati Tags)

Social bookmarks full-text search (there isn’t one)
Social book marks tags search (tagbert, Tagcentral, Technorati Tags [kind of]…although none of these present mixed results in the main body)

A combination of both.

[ADDED 27/09/05: Guten Tag]
[ADDED 12/10/05: gada.be]
[ADDED 14/10/05: Wink]