Library clips

Don’t know about the name, seems a bit familiar, anyway Diggol is in the realm of Findory, Spotback and Rojo…based on your click behaviour it learns your tastes and delivers personalised news, at the same time your interactivity adds to the community page.

It’s in alpha at the moment, so I see this service developing into something special, as I’m already impressed.

The great thing about Diggol is that you can enter your OPML, just like at TailRank, Findory, and megite…and that’s not all it has is common with these memetrackers.

Discovered Topics (Word Bursts)

What I like about it is that your news stream is organised by topic, must be something similar to how Zoom Clouds works…Rojo has a great relevancy feature, but imagine reading Rojo based on word burst topics (well now you can, just plug in your Rojo OPML into ZoomClouds or even PersonalBee to try…actually plug it into Diggol).

At the moment in Diggol, each item has word burst topics, or you can view the word burst list (discovered topics) on the sidebar…these are based on your OPML (this would look could in a cloud view).

What Diggol doesn’t have is a tag cloud derived by user tags from the feeds in your OPML…I guess if you imported your OPML or your Rojo RSS into SuprGlu or MySyndicaat, this would be a way to read your subscriptions by user topics, or for each item discover other items by clicking on an author tag.
I posted on this a while ago…reading your feed set by wordbursts or user tags.

In respect to what I call word bursts (discovered topics), I’ve been told it does more than simple keyword extraction…from an email by the developer (Ping):

“We do a lot of heavy lifting natural language processing, perform part-of-speech analysis, measure the informativeness of the words and phrase to make sure the extracted Topics carry enough meaning for a user to decide what to pursue…

In addition, our Topics are not ranked by how many times they appear like the tag clouds do. Instead, they are ranked by importance, measured by what posts they are in, how important that feed is, how many users read posts containing this topic, how much the individual user (after login) is interested in this topic (from his past reading behavior of similar topics), among other things.”

More from the FAQ:

“The list of topics after a post shows the clusters of posts that cover same or similar concepts. They also serve as auto-tags of the post.”

Related Posts (Memetracker)

This notion of discovered topics is the idea of a memetracker, but not quite the same

…so perhaps it really is a personalised reader crossed with a memetracker, the difference being it also includes clicking behaviour (this is the personalised reader part), and it shows all posts from your OPML (whereas TailRank and megite only show some posts from your OPML whilst recommending posts)…Diggol keeps recommended posts in a different section.

I’ve just noticed that some items have “Related Items”…so now it really is developing into a Memetracker
…so now you can find similar stuff via “Discovered Topics” and related stuff via “Related items”.

One thing it doesn’t do as of yet is have a section for aggregated outgoing links from all posts, like Chuquet…actually Rojo once had this feature in a section called Recommended Links (use CTRL-F to find it).

Also another thing I’ve mentioned before is threading…if a read or unread post in my RSS Reader has a hyperlink in the body of the post that is the same as the hyperlink in the post I’m reading, then thread them together.
ie. if a post I haven’t read yet has the same link in the body of the post as the post I’m currently reading, thread it, then I can click on this unread post and read it now.

Relevance

Besides the tracking your reading behaviour eg. clicks, comments, voting, tags…there is a whole lot more going on…again from my email correspondence with Ping:

“Before login, the posts are ranked based on several factors: the authority of their feeds, the topics they cover, how the topics are related to and reinforced by other topics (an iterative process, much like the google pagerank, but uses topic relations instead of hyperlinks), users’ collective interaction with the feeds and similar or related topics in the past…

After a user logs in, the user’s own past interaction (time decayed) with the feeds and similar or related topics in the past are weighted much higher in ranking the posts…”

Back to it…

Import your OPML, or choose a category from the Diggol feed set, choose them all if you want the whole Diggol feed set.

It seems you can even organise your feed set into categories…some people have their subscriptions already organised into folders, maybe this is something that could be done when you import your OPML.

You can even filter keywords in your feedset, these are called My Topics (sometimes known as “smart feeds”).

You can also make your Diggol public/private/share just with friends.

So lets get this right:

My Feeds - your OPML (this appears in the My Topics section)
Categories - organising your OPML (I think?)
My Topics - filtered keywords in your OPML
Discovered Topics - wordbursts…from your OPML
My Tags - items you have saved
Related items - memetracker feature (clustering)

I really like how you can view the items from aspects of the attention data…
My Tags/Comments
Already Read
Commented
Real-time Recommendations

Then at the community level:
Hot Topics
Hot Tags
Most Voted
Most Read
Top Voters/Taggers
Public version
- Social aggregation (all your attention data is aggregated for the community version)
NOTE: What I mean by attention data is your OPML and how you read it.
I wonder if they will promote some users accounts, I’d like to see peoples OPML’s through the Diggol attention filter.

More…
Topic map
Trend graph
Search

NOTE: my OPML didn’t comply with the template for importing so I got some help…here it is.

RSS Reading

Read by river of news
- My feeds
- Categories you created
- My Topics

Read by feed
- but you have to click on the feeds name under the title of a post
…I’d like this listed on the sidebar, with a mark read/unread feature (Traditional RSS Reader)

Each post

Discovered Topics
Categories
Tag
Comment
Email
Vote
Related posts

What I’m seeing here is a tool that combines the best features of Digg, Zoom Clouds, Findory, Spotback, Rojo, TailRank, and Megite (both at the personal and public level).

In the end to use Diggol, I’m going to have to give up Rojo, and I just can’t do that, I need to keep track of what I have and haven’t read (mark read/unread)…plus I like having a subscription list in the left pane.

I like the idea of TailRank and Megite as they seem complementary to Rojo, whereas the personalised reading behaviour readers like Findory, Spotback, and Diggol are more of an alternative…again including a read/unread feature would make it a very competitive alternative.

Although I can one day see Rojo having a different view in addtion to date, and relevancy, and that is “meme”…basically incorporating a memetracker view.
eg. view just a selection of posts from your OPML, with recommended post thrown in to the mix.

And of course for each item, showing related items (from the blogosphere or just from your OPML, or from your contacts OPML)…this could also be a feature in the date and relevancy view.

While I’m at it, bring back the aggregated outgoing links section…and as mentioned before, view posts by wordbursts, author tags, and the beloved threading feature.

Congratulations to Diggol for incorporating so many attention features into one service.