Library clips

Every type of search on Furl, generates an RSS search query feed…also bookmarks can be saved in multiple folders, have assigned multiple keywords (I think) or a combination of both folders and keywords.

It allows to show individual archives of people who saved the same site, from these we can view the difference in filing habits.

A query searches the free full-text of the archive…you can restrict search to one folder or, search with the keyword prefix you may have assigned when you first bookmarked the webpage, eg. keyword:cats

As far as I know a query does not search or browse the label or name of the folder (there isn’t a specific search field called labels or folders or tags) in all furl accounts…well you can browse to a certain extent, if you go to the "latest headlines" section there is a running list of popular folder names. Clicking on one of these will show you all webpages labelled with this folder name with a person’s name next to each entry and the number of other people who also saved this item. On this note, I realise it lacks to show what other label or folder they may have applied to this item (remember you can give an item multiple labels/folders)..this is a handy feature in del.icio.us.

With Furl you actually have to click into the person’s account in order to see other labels they may have given the same item, this is a bit cumbersome when compared to the inducing social browsing power of del.icio.us(explicitly shows all tags a person applied to this URL).

RANDOM EXAMPLE

Furl

In search for a virtual settlement: [by draggin (and 3 others) at 19:27:53 GMT on Feb 14]

del.icio.us

You can always goto to the URL in the address bar and guess a folder name that may be in use…the example below is the URL for the folder "tools", replace the word "tools" with another term, if you fail to get results this means no one in Furl has ever applied that folder name…same goes with the tag blog in the del.icio.us URL.

http://www.furl.net/furled.jsp?topic=tools

http://del.icio.us/tag/blog

In this instance a search field by folder would be great (that is searching folder names), the results could be returned alphabetically, if  your search term isn’t there it could show where it would live if it were to be used in the future. This also applies to del.icio.us, searching for tag names, instead of just browsing or searching in the title field (description) - not sure if it searches in the extended field as well.

I wonder if you can search in all Furl accounts by the keyword prefix, not sure everyone uses this function, I know I don’t at the moment.

Related

When you click on a tag from the homepage on del.icio.us it shows related tags which Furl doesn’t do, and when you click on the link for a given URL, for example "and 44 other people" it shows common tags..del.icio.us is great at this, it keeps offering you tags to browse. Although the good thing about Furl is it explicitly shows a summary of common interest URL’s, "people who furled this also furled…" which is a handy feature because it’s so accessible.

For me to follow a topic like "folksonomy" on del.icio.us, I would browse my tag, see related tags on offer (such as distributed classification, social, tagging, free tagging, etc), go into accounts to see even more tags used for similar content URLS, then I could put these all into the "inbox" and read it from there or put the RSS feed into my RSS reader. To do this on Furl is more cumbersome as I would have to actually go into other peoples accounts just to see what folders they use to describe content I’m interested in.

If you don’t want to mix your "inbox" (del.icio.us) or "my headlines" (Furl) with other tags or accounts you follow you could always use Blogdigger to generate a compilation RSS feed.

Duplicates

I wonder if del.icio.us (same goes with Furl) can take out duplicate URL’s when reading in your "inbox?"…can my computer know if a URL has been in my "inbox" before (I know it changes colur if I’ve previously clicked on it). If a URL has appeared in my "inbox" before maybe it could be ranked to the bottom of the list and be of a different colour, if I have clicked on it a different colour again, and if I have already saved it in my del.icio.us account a different colour again. The only problem is that these visual alterations would have to carry over to work in my RSS reader.

The reason I would like this feature is that if you subscribe to a tag, such as RSS, and don’t read it for a few days you accumulate hundreds of unread posts (a filtering mechanism by past browsing history would help alleviate information overload from seeing the same sites again and again)…I guess I want to see sites that are new to me personally, instead of what’s recently been tagged.

Personal keyword index anyone?

Now what I can’t get my head around is the blurring of features between these two social bookmarking utlities.

The folders in Furl are similar to del.icio.us tags, but it seems that Furl has two layers of tags, one for folder and one for keyword (but they don’t neccessarily have to have a relationship)…also to note is that keywords haven’t got the socially functional features of folders - they aren’t explicitly visible, unless you know about them.

Users can treat labels/tags/folders in a way that suits their cognitive thinking. In Furl a folder could represent a category/class heading and keywords as subject terms. In this case I don’t have to have folders like the following: RSS-tools, RSS-search, RSS-readers…I could instead have one folder called "RSS" and within this folder have the following keywords: tools, search, readers…

A definite requirement for Furl is a personal keyword index, so I can see my list of keywords I have used in the past (this way I can view my keyword list before applying a new keyword - keeps things consistent)…I’d like this to be at the folder level as well. You could have a fielded drop down menu showing a keyword index for the total of all folders in your account and a keyword index just within a folder. When I last emailed Furl, the response was that this is a great idea, but that was a while ago…

My headlines RSS

…they also took my suggestion that an RSS feed for "my headlines" was also a good idea, so you can read "my headlines"in your favourite RSS reader instead of within Furl (del.icio.us has an RSS feed for the "inbox"). Sure you could put all the individual RSS feeds of Furl accounts you subscribe to in your reader, but this gets too much when you know it could be so much easier…I guess you could always use Blogdigger to make one feed.

Retrieval

Great thing about Furl is that it is your own mini-search engine that searches the full-text of your version of the web, also searches all Furl accounts, and the web at large…del.icio.us can’t do this of yet, but it has a unique combing (+) feature that allows you to retrieve items that are assigned with more than one tag, Furl doesn’t do this for retrieval it only does this at the initial bookmarking stage.

Popularity

People start tagging items with existing popular tags, (general/simple tags are usually popular), whereas there may be a more appropriate or specific tag (which you may have to invent or one that is simply not popular).

…maybe we need major descriptors and minor descriptors

Personal vs. Social

This leads me to think of the style of tagging I would use when bookmarking information in order to retrieve it successfully - would it be according to my personal style compared to applying tags with labels that are more general, in turn more socially viable, ie. are my tag headings for my personal benefit in organising and retrieving, or for the benefit of the social network. This is the compromise, as I like to use tags that are suited to my intuitive style, but then I’d like others to use tags that are general enough for me to come across (if everyone used complicated tags it wouldn’t be as an efficient sharing system).

EXAMPLE:

Personalised version

LIS_rsch          LIS_edu          LIS_km            LIS_rss          LIS_tags          LIS_applications

Blog_rsch        Blog_edu        Blog_km          Blog_rss         Blog_tags        Blog_applications

Sociably friendly version

LIS                    Blog                     Education           Applications

Research          KM                      RSS                      Tags

In comparison the personalised version would accumulate a longer list of tags over time, scrolling through a large list to find the right tag…although the tags are more precise, if they are too precise you may need to invent a new tag to add, increasing your already large list. The sociably friendly version would have a smaller and more manageable list, but it takes more thought at the time of tagging using the correct multiple tags in order to build the search (+) when retrieving later on.

But then, for example, the tag "Research" really is misleading, to me it is research about blogs, whereas to others "Research" may be about geology, dogs, food, etc..(they mean something different within different domains, and within different accounts)

When there are multidisciplines and the content starts expanding, a folksonomy (or a controlled vocabularly to a certain extent) both don’t scale as well, and search engine free-text ranking takes over as the preferred method. Added to this are also the well discussed problems of spam in tags, synonym control and the much more… (tag stemming is a great start as a clean up tool, if only this could become more intelligent).

Channels, domains, disciplines, directories…

To accomodate multiple interests it would be good if your social bookmarking account could be split into channels of interest, a monitoring service, WatchThatPage, offers this function which I find very useful, it saves me registering several accounts.

If we have a future problem of the tags losing value as the service grows large with content, maybe we do need channels or discipline specific services, such as Connotea (social bookmarking for Science). In this way the bookmarks in the social network have a common interest and won’t be mixed up with other irrelevant interests such as knitting. Other bookmark services promote their service with a specific value, such as CiteULike (social bookmarking to share properly formatted references or citation/abstracts).

Maybe general social bookmarking services, such as del.icio.us or Furl, could have a directory like Yahoo, with a check box next to each domain of interest. When you register you check your boxes appropriate for the content that you will bookmark, or you could just check them all…or even a social bookmarking services regitry or network…this is getting over my head…

For now…

In the end I find folksomies of great value in providing me with my own searchable version of the web, and as an effective way to find relevant things of interest. Without del.icio.us or Furl I wouldn’t have come across all the great information, and all the insightful people…they are great social tools…I have a new found love for the word "Serendipity."

…it’s also great that the social aspect of these tools continue to generate ideas for the benefit of its own future, it spores discussion for its own betterment.

NOTE: Since I like to use both Furl and del.icio.us as they both have their unique features I’d like to thank Alan Levine (CogDogBlog) for making a most handy bookmarklet.