Years after the untimely demise of Share Your OPML and available now, unlike the super feature rich but frozen in private beta service Toluu, Jonathan Christopher's SuggestRSS is easily worth the minimal effort it will take you to try out.

To use SuggestRSS, export your reading list from your feed reader and you'll get an OPML file on your desktop. Upload that file to the SuggestRSS website and it will do a simple comparison. People who like what you like also like these things you don't have. Simple as that. Amit Agarwal found and blogged about it first, among people we know.

You'll get some duplicates with the system, the sample size is small, it's dominated by tech feeds and not good for much yet with other niches. But you can come back to the URL of your recommendations later and find new suggestions!

That's it. There's nothing more to it. Now if you'll excuse me, I can't believe I had never thought to subscribe to the feed for Boston.com's Big Picture. It's really incredible photography and I appreciate SuggestRSS prodding me to subscribe.

During the last few months, BlogRize's founder Jesse Spaulding has been working on a major redesign of the site, which he is rolling out today. The new design features an enhanced voting system, updated ranking algorithms, and a lot of updates to the user interface that make using the site a lot easier and more fun.

BlogRize creates a community of readers around specific blogs and then generates an individualized Digg-like site for each community, where the ranking of the stories depends on the recommendations and votes of other group members (among other things). While you can submit stories to BlogRize directly, the main conduit through which users add stories to the system is through recommendations in Google Reader.

User Interface

A lot of the user interface updates to BlogRize are quite useful. You can now, for example, toggle story previews and see in-line comments from your friends right on your BlogRize homepage. This new version also allows you to quickly mark a story as read by simply clicking on the white space around the story.

On nice addition to BlogRize's feature set is its ability to find your profiles and activity on other sites through Google's Social Graph API. This makes importing your profiles a lot easier.

Most importantly, however, Spaulding has streamlined the voting system, which was one of our few complaints about the earlier version. Unlike Digg or Reddit, where you can only vote a story up or down, BlogRize allows you to mark a story as 'interesting,' 'funny,' 'disagree,' 'seen this already,' or 'inaccurate.' BlogRize also looks at links to stories from other blogs and takes these into account when it ranks its stories as well.

Join the RWW BlogRize Community

Jesse told us that his focus while developing and redesigning the service was on giving bloggers an opportunity to create and promote their own blog communities, and after this redesign, BlogRize has become an every better place for blog readers to get together in a relatively small but focused group. Thanks to this focus and the self-selection of the group members, the recommendations are always spot-on.

If you want to join the community of RWW readers on BlogRize, just click here and sign up for the service.

While it did take a bit of training before the application fully recognized our preferences and before it returned really good results, the overall results were very impressive.

So how does it work? When you first sign up, my6sense can import your RSS feeds (from Google Reader, Netvibes, MyYahoo, or from a standard OPML file). After that, all you have to do is read your feeds through the web app and my6sense will automatically learn from your reading behavior (my6sense calls this "digital intuition"). You can also explicitly give a 'thumbs up' or 'thumbs down' on any post.

My6sense is also a very capable RSS reader in its own right, with the built-in ability to share items on Twitter, Facebook, FriendFeed, and LinkedIn, as well as by email. One very nice feature of the app is that it also allows you to see comments on posts from supported blogging platforms.

Currently, my6sense is only available as a web app, but the company expects to have a native iPhone application in the App Store by the time the service launches its public beta early next year.

Training

On a mobile device, good recommendations can be quite a time saver. Having to scroll through numerous feeds can quickly become frustratingly slow. In our tests, it took a day or two of regular usage before my6sense started to come back with really good recommendations. As Kristie Wells, my6sense's VP of Marketing and Community Relations told us, the company is working on providing a better out-of-the-box experience that will learn faster, but given our positive experience, training the software off and on for a day or two is well worth it.

Attention

We also talked to the two co-founders of my6sense, Barak Hachamov and Avinoam Rubinstain, earlier last month and we asked them if they were going to support any of the open attention profiles like the APML, which would make it possible for users to take their attention profile and transfer it to other news readers. For now, however, the company says that it is focused on providing a good user experience for its alpha testers, though they didn't rule out support for the APML in the future.

Invites

Our experience with my6sense has been very positive, but you don't have to take our word for it. While the service is still in private alpha, they have provided us with 500 invites for our readers. Just click here to claim yours and let us know if it worked as well for you as it did for us.

Web Trends

RWW Live: Health 2.0

We have been tracking the so-called 'health 2.0' trend for some time now. We've covered the top health web apps, the trends to watch in health 2.0, and the latest industry stats. And we published a Health 2.0 update this week (see below). So in this week's episode of RWW Live, the ReadWriteWeb authors get together with a number of industry experts to discuss how the Web is changing health care.

Download MP3

Health 2.0 Through the Eyes of a Diabetic - One Year Later

ReadWriteWeb founder/editor Richard MacManus: One year ago, I discovered that I had contracted Type 1 Diabetes. I was 36 at that point and it's relatively rare for someone of my age to suddenly get Type 1 Diabetes - indeed they used to call this form of diabetes "juvenile diabetes", because it mostly occurs in children. So it was quite a shock to discover that I had it! Immediately I looked to the Web to find out all I could about this condition. I discovered a thriving community of 'health 2.0' apps and social networks, which I then wrote about in this blog.

As it's now a year later, I thought it'd be good to review health 2.0. What has changed in web-based health services over the past year? And indeed what web tools do I use to help me manage diabetes?

Nine Recommendation Tools We Wish We Had

...And The Best Substitutes We've Come Up With So Far

There's so much content online every day that it's totally overwhelming. That's where good recommendation technologies and media outlets come in handy. As a blog that seeks to share the most interesting web technology and trends with readers, automated help with the discovery process is of great interest to us. In this post, we discuss some tools we wish we had and the closest makeshift substitutes we've been able to come up with. Maybe you'll find some of them useful or have even better recommendations to offer us and other readers.

2009 Tips for Big Web Companies

2009 is approaching quickly, and the consensus is that it's going to be a really tough year. The US financial crisis is triggering a global recession. Yet, a crisis is also a time full of hope. It is a time to re-think, re-tool, and get ready for the next upswing.

For big Internet companies, 2009 is going to be a very bad year for sure. Advertising profits are going to plunge, and consumers will spend less money overall, particularly on the web. There is little that can be done to change that. But what big companies can do is invest in innovation and killer moves that will bear fruit in the years to follow. Here is what we think would be cool for various big web companies to do in 2009.

SEE MORE WEB TRENDS COVERAGE IN OUR TRENDS CATEGORY

A Word from Our Sponsors

We'd like to thank ReadWriteWeb's sponsors, without whom we couldn't bring you all these stories every week!

• Mindmeister is an online mind mapping tool.
• Pro Opinion allows you to get paid to take surveys.
• WildApricot is Membership Management Software.
• Crowd Science gives you detailed visitor demographics.
• Glue enables you to connect with friends while browsing.
• Strands provides real-time recommendations of products.
• Direct Media Exchange is a solution for managing ad networks.
• Rackspace provides dedicated server hosting.
• VisualCV lets you stand out from the crowd when job-hunting.
• Quintura is a visual-based search engine.
• Eurekster is a custom social search portal.
• MediaTemple provides hosting for RWW.
• SixApart provides our publishing software MT4.

RWW Jobwire

Who's Getting Hired in Tech? Last Week's Jobwire Graphs

We all know the economy is in shambles and there are massive layoffs across most industries, including technology - but there are still new hires happening in tech and new media! Who's getting hired? That's what we're tracking at our blog Jobwire - the stories of lucky people with exciting new jobs.

Who's getting those jobs? What kinds of positions are being filled and in what sectors? Check out our first set of charts below from last week's aggregate activity on Jobwire to find out the answer to those questions.

SUBSCRIBE TO READWRITEWEB'S JOBWIRE FOR THE LATEST NEWS ON JOB HIRES IN TECH

Web Products

Google SearchWiki is Not a Wiki

Late last week Google unveiled the first major change to its search interface since the introduction of multi-media ("Universal") results into the search results page. They called it Google SearchWiki.

It's a big deal, it's awkward, it's frightening, it's brave, it's already both loved and hated - but it's not a wiki. As Ward Cunningham, the man who invented wikis, told us in our initial coverage of SearchWiki, "Collectively editing thoughts is what leads to the unique wiki behavior..." Days into the experiment it's clear that this feature is more like a forum, and it's not a particularly well architected one at that.

Your Favorite Mobile Web Apps & Sites

One year ago we ran a contest asking you to tell us your favorite Mobile Web apps. From the resulting comments there were 5 Mobile Web apps that clearly stood out, with multiple mentions: Gmail Java app for mobile phone, Google Maps for Mobile, Opera Mini, Fring, Shozu.

Well, a lot has changed in the Mobile Web application world since then. The Apple App Store launched in July '08, prompting a wave of new third party iPhone apps. And we've seen innovation from Apple's mobile phone competitors: Google's Android (which has multiple app stores), Nokia, and Blackberry, and others. So what are your favorite Mobile Web apps and sites circa November 2008? The ReadWriteWeb authors listed their faves, plus we polled our friends in Twitter (subscribe to our Twitter account @rww).

Firefox China Edition: Everything a Local Browser Should Be

Did you know that the way you surf the internet may be influenced by your culture? In the U.S. and Europe, web surfers are leaning forward, one hand on the mouse and the other on the keyboard, typing and mousing equally. In China, however, the process is much different. Web surfers there tend to lean back from the monitor while keeping one hand on the mouse, the other hand dangling. The keyboard is used much less frequently as much of the navigation is done with clicks instead.

YouTube: More to Love

As the prices of professional quality video equipment continue to drop and the number of people with high-speed internet connections continues to increase, online video sites have been scrambling to keep up with their users' desires to deliver higher quality content to their viewers.

YouTube is no different. This week, they announced the latest enhancement to the YouTube platform - a widescreen video format across the site - which they hope will provide users with "a cleaner, more powerful viewing experience."

SEE MORE WEB PRODUCTS COVERAGE IN OUR PRODUCTS CATEGORY

RWW Enterprise Channel

Is SaaS Cheaper Than Licensed Software?

Most people quickly answer this question in the affirmative. We certainly do. However, there are people out there who aren't sure. They look at the monthly cost of a SaaS application and compare it to the equivalent licensed product over an extended period of time. Given enough time, you will eventually hit a point when the SaaS product appears to be more expensive. In this post we looked at it from the perspective of the total cost of ownership (TCO).

Email us if you're interested in writing for ReadWriteWeb's Enterprise Channel.

SEE MORE ENTERPRISE COVERAGE IN OUR ENTERPRISE CHANNEL

That's a wrap for another week! Enjoy your weekend everyone.

Idiomag is a company we love to tell people about - it's one of the most awesome ideas we've seen in a long time. The personalized music magazine site relaunched today with 4 times more content than ever before and a more traditional, link-intensive page design. Unfortunately, this great idea has serious problems in implementation and today's redesign doesn't appear to have solved those problems.

Idiomag looks at the color scheme of the video or photos accompanying an article and bases the whole page aesthetics on a complimentary set of colors. The service learns from your feedback and offers a new set of recommendations ever day.

We love the idea, but two years after Idiomag launched, we still return to the site begrudgingly today. We visit it at every opportunity, appreciating the music recommendations and hoping the rest of the site will get better.

Content is Hard

Why is it so unpleasant to go to Idiomag? Mostly because the syndicated text content from music blogs is awful. Most of it is really poorly written and the presentation is even worse. The Flash display of text from feeds misses punctuation, inserts spaces between letters in the middle of words - it's absolutely unreadable.

We keep hoping that the next article served up will be free of overdrawn adjective soup or abrupt beginnings and topic changes. No such luck.

There's obviously a world full of crappy music blogs out there, and we appreciate the MP3s they post, but Idiomag's parsing for quality content isn't working. The blog posts are placed very prominently on every page of the site and the whole user experience suffers as a result.

There are a lot of moving parts in the UI that make it unpleasant to use, as well. The first paragraph of every article is repeated as a "pull quote" at the top of the page, meaning you've got to skip the 2nd paragraph on every page because it's the same as the first. There's an independent collections of MP3s playing on all the pages, luckily they stop when a video loads, but they are no longer tied to the same artists as the articles and it appears that every article now has videos instead of photos.

We didn't see any of the celebrated new content from partners like Billboard and Rolling Stone. As part of the relaunch today, Idiomag says it has quadrupled the number of articles it can offer - up to 11,000 per month.

Give it a try for yourself, it's a great idea and maybe it will serve you better than it has us.

For fans of Muxtape, the new PodLists feature may hold some appeal - the only difference is that instead of searching through various online mixtapes identified only by username (you should at least be using Muxfind for that!), you'll have daily access to a fresh set of what are, essentially, online mixes identified by theme or genre instead.

As you listen to the PodLists, you can use SeeqPod's built-in features to remove the tracks you don't like and save the rest to a playlist of your own.

At the moment, the PodLists aren't customized to a particular user's tastes, but the SeeqPod team does look at the most popular search queries by topic or artist, and, based on the data found, they then create a list that suits that particular demographic. However, sometime this summer individual users will be able to search for lists by username, list name, or date - which will make the service more Muxtape-like, considering that users could then create playlists that others could discover.

For anyone in search of more ways to discover new music, this service is worth a look...as long as they keep their recommendations fresh.

For more information on SeeqPod, you may be interested in checking out our recent Alt Search Engines podcast which featured LaurieAnne (LA) Lassek of Seeqpod and Aza Raskin of Songaza (who partnered with SeeqPod this year to offer more music).

Anton Talks About The Digg Recommendation Engine from Kevin Rose on Vimeo.

Not all the users have these features enabled yet, but those of you who do can check by going to upcoming and checking for a red BETA label. The new upcoming system has three ways to sort it and the third option in the list, Most Diggs, is the one you're used to seeing, where all stories are presented in order of decreasing Diggs.

The first new option, Most Matches, looks at your history of Digging, compares it with other community members, and shows the stories in order of number of matches. In the case of the first story, you see the expanded view of the 'Recommendations via' list, and in the case of the second story, you see it in the compact version, not showing the user names and percentages, rather only the total number. For example:

The second new option, Most Recent, shows you the stories recommended by community members compatible with you, in reverse chronological order. You can also see why a story was recommended to you (via user name and percentage of compatibility with that user).

A new section in the sidebar, entitled 'Diggers Like You' shows you Diggers that are most like you in their Digging and submitting habits.

And finally, you can click on a user and compare exactly how much you overlap. In the screenshot you see below you can see the overlap between my profile and thediggboss's profile. In total we had 3864 overlapping Diggs in the past 30 days, which means our compatibility score over all our Digg activity is 38%.

Overall the design is great and there is a decent feature set. As far as what it is designed to do, it seems to function well. At the same time however, whether the engine will help content submitted by a fairly obscure user, remains to be seen. In the beginning, all your compatibilities are going to be with the people that you have been Digging and the people that have been Digging you back, i.e. your friends. It will require widespread use of the feature 'Diggers Like You' to help more obscure submissions travel to a lot of people.

It is also important to note that the recommendation engine will be a boon to advertisers as well (and of course Digg). By sending the most relevant links to the most relevant people, you can also send the most relevant advertisements to the right people (and ensure high-quality clickthroughs). Users get good content and related, hopefully non-intrusive ads, advertisers get the right potential customers, and Digg gets the money.

This is a guest post by Muhammad Saleem, a social media consultant and a top-ranked community member on multiple social news sites. You can follow Muhammad on Twitter.

Special thanks to thediggboss for providing the images.

Finding the right apartment to live in can be a daunting and time consuming task. Introin tries to make this process easier by creating a platform where prospective renters can communicate with current residents. Current residents can earn referral bonuses from their rental complex when they convince a prospective renter to move to their complex. Introit also gives renters the opportunity to advertise their own apartments for subleasing.

Introin currently only has listings available for the United States, but there, it does seem to have a relatively complete list of apartment complexes. What makes the site a lot less useful, though, is the fact that the search functions are extremely basic, to say the least. Other sites focused on the rental market like ForRent.com or Rent.com allow you to filter your search by everything from available fireplaces, to pools, and pet policies. Even Apartmentratings.com, which mostly focuses on reviews, allows for searching by number of beds and baths, as well as maximum rent.

Introin, on the other hand, only features a location based search and an option to search by name. That's it. After the search is completed, you are presented with an alphabetical list that can only be sorted by number of referrals. The information Introin displays about apartment complexes is about as spare as its search functions and doesn't go beyond name, address, and phone number. There are no photos, no links to a complex's web site, not even an indication about the cost of renting there. There is a button that says 'show details,' but it's inactive.

Connecting Renters

Even though the search presents a pretty bare list of information, this is the central hub for renters to connect. You can add your own information here to be listed as a referral and you can list your apartment as being available for sublease. At least in the sublease form, you can start entering some information about your place. You can not, however, upload any pictures.

Once a connection between renters has been made, the communication between them is then handled on the site on a message board. If there are no subleases or referrers available yet, you can set an alert and you will get an email once somebody else signs up for referring or subleasing.

Conflict of Interest?

If current renters, though, are mostly motivated by referral bonuses, the question remains if their opinions can be fully trusted. Many complexes offer bonuses close to $300 or a certain number of rent-free weeks. Introin doesn't feature any reputation system. This is problematic in a system that has a conflict of interest built into it by default.

What if Nobody Shows Up?

The problem for Introin right now isn't this potential conflict of interest, though. The problem Introin faces right now is one that many community sites face in its early stages: it's empty. In all my searches, I have yet to come across a single apartment available for sub-lease or a users who registered as a referral.

Introin is probably most useful in areas around college campuses where people move in and out on a regular basis and somebody is always looking to sub-lease apartments for the semester breaks or after dropping out of school. Here, however, Introin also faces stiff competition from long established local sites and, of course, Craigslist.

Verdict

Introin is build upon a very interesting idea. If the developers tackle some of the problems outlined above and start adding more information to the system so that the search function becomes more useful, then this service would have potential. Right now, there is simply a lack of activity on the site and the fact that the search is so limited reduces its usefulness to the point where it would be hard to refer anybody to the site with a clean conscience.

How to Use it

InSuggest for bookmarks is very simple. You enter your Del.ico.us username, it looks at your archive of bookmarks and then recommends other similar pages you might like to bookmark. You can filter by one or multiple tags you've used. Up to 20 tag filter options are provided but you can enter any tag you've used in your account.

You can run anyone's Del.icio.us username through InSuggest and get recommendations, not just your own.

The interface is very nice, it's one of the best uses of Ma.gnolia's Thumbshot.org that I've seen yet, and the whole thing feels fairly smooth. In fact, it almost feels too smooth. The recommendations come to you very, very quickly. No where on the site, or in response to our email inquiry so far, can we find an explanation of how it works.

Despite that, it does seem to work well. There are a limited number of ways to parse Del.ico.us data, though, and we wouldn't be surprised if there's just a touch of caching going on. It's hard to say, but the end result is good. InSuggest developer Dennis Gustafsson was elected "Engineering Hero" by The Swedish Association of Graduate Engineers last year, according to search blog Pandia. So someone's seen behind Gustafsson's work and liked it.

Continued below.

Our Recommendations

Beyond some clarity around process and perhaps basic instructions on use, there are a few other things we'd like to see from InSuggest. The first is a feed for future recommendations. The display is in Javascript so we haven't been able to scrape it yet. We'd also love to see a Greasemonkey script for displaying InSuggest recommendations on top of the Del.icio.us bookmarking popup, archive page and item pages.

Other features that would be nice would be the option to input a Ma.gnolia username instead of just del.icio.us, tooltips to display full item titles that are too long for the basic display and the ability to exclude particular domains from future recommendations. Some sort of user feedback to inform recommendations should be doable.

Finally, the biggest fish in the pond when it comes to Del.icio.us recommendations is user recommendations, not just item level ones. We'd like to see other users be recommended, ideally with those who tend to find items of interest earliest privileged on the list.

That may be too much to ask for, though. It's hard to say. Feeds and user recommendations are the kinds of gifts that keep on giving, though, and are far more compelling than one-off recommendations.

For now, though, we think InSuggest for bookmarks is worth checking out. You could very well discover some things there that are just the kind of thing you've been looking for.

Strands currently has 150 employees and it sees 2008 as "our year to market". Strands told ReadWriteWeb that it plans to ultimately apply its recommendation technologies to 3 areas:

1. Business solutions - helping people find content in a client site
2. Social media - Strands says it will soon launch services "to help people discover things they might like, based on their online behavior" and "help them make sense of all their dispersed social media activities"
3. Personal finance - helping people consolidate their dispersed financial data, and helping them find new ways to save money and invest.

Strands sees these three areas as having "strong personalization challenges."

moneyStrands + Expensr

The personal finance part is what is being announced today. Currently in private beta testing, moneyStrands is "an online money management solution that allows users to aggregate their online financial information in one place, providing them with an instant snapshot of all their finances." Similar products on the market today include Wesabe, Mint and a kiwi startup with global ambitions called Xero. moneyStrands is employing recommendations technologies, such as enabling users to anonymously compare themselves to others with similar traits - e.g. demographics.

Expensr, the app Strands has acquired, is a free online application that combines social networking with financial management. There is no word yet on whether Expensr will be integrated into moneyStrands; and if so how. As of now it will continue to run as a separate service.

Conclusion

As we've noted before, we at ReadWriteWeb are following the trend of recommendations closely - it was one of the 5 major trends we outlined in our toolkit for 2008 and was featured in my Media08 presentation Web Tech Trends for 2008 and Beyond. Strands continues to pique our interest, but let us know in the comments what other recommendation startups we should be looking at too.

The DataPortability project taps into the strong conviction, engendered by the do-it-yourself nature of the web 2.0 movement, that individuals should "have control over their data by determining how they can use it and who can use it". This extends to an inherent belief that that it has not been a lack of effective technology that has held back this new culture of open data sharing, but rather business models that have been over-reliant on laying a proprietary claim to some portion of that data.

Taste sharing is a DataPortability use case

We express our online tastes any time we make a choice between the various alternatives available to us. Some of our choices may be characterized by the number of times we select each option when repeatedly confronted with the same choice - for example the number of songs of each genre we play when we select music. Other choices may be expressed subjectively, such as assigning one to five stars to movies we watch, when we are asked to rate our preferences for the different alternatives. In yet other cases, we may in effect be giving estimates for the number of times we would expect to select each alternative, such as when we are asked if we are likely to buy a product or vote for something. Virtually any online experience we have includes one or more instances in which we make conscious choices reflecting our preferences.

For the more theoretically inclined amongst us, we can view a choice as somewhat analogous to a random experiment and our relative preferences as measures of the different possible outcomes of the experiment. The collection of such experiments that we participate in as a matter of course in our web experiences paints a vivid picture of our taste. For the more pragmatic: each time we make choices, we generate data which empirically describes our preferences. This is data that can be encapsulated and shared just like any other picture, blog post, video, or other piece of online content that we create; and which the DataPortability project is focused on.

A few ideas for open taste sharing

As a DataPortability use case, open taste sharing embodies and embraces the culture shift that the Web 2.0 movement represents. With regard to data ownership, the DataPortability concept has even more succinct expression: our tastes should be ours to share, or not. This puts the user in control of their online experience, so they can set the boundaries of how much they want to share and with whom. Similarly, there is no need to invent new or proprietary technologies to simply identify, encapsulate, and share taste-related data.

A little thought by websites about how to identify and summarize our relative preferences on their site is enough to do the job - along with OpenID, OAuth, and a little task-specific XML for markup. However clearly this kind of data sharing also raises new privacy concerns, which is part of the work-in-progress for the entire DataPortability project.

Image by MyStrands

Perhaps the most interesting challenge lies in educating businesses to thoroughly and thoughtfully examine their current ideas about user data, so they can better understand and enthusiastically embrace The Web 2.0 Golden Rule: "Do for other web experiences providers as they would do for you - under our control as the owners of our taste data - and the blessings of networks effects for taste data shall be yours."

This is a guest post by Dr. Rick Hangartner, MyStrands Chief Scientist. You can follow Dr. Hangartner on the MyStrands blog.

So while we don't know how each service works, surely we can figure out which works better. To that end, we put them head-to-head and came up with our opinion.

DiggFilter

The point of DiggFilter is to suggest "fresh content," according to its founder. "There are many, many stories submitted to Digg on a daily basis, more than anyone could ever sift through. Most of these stories never make it to the front page, and as such are undiscovered by the majority of the people. Hopefully the tool will help users find these hidden gems that the majority of the community moderated so they couldn't be promoted, but may be of interest to you," Mulvyava told us.

But after entering my name, DiggFilter told me that I was "a tough one" and warned me that my suggestions "may not be accurate." That didn't really inspire confidence in the service. However, the suggestions were actually pretty good -- it was mostly stuff I would digg. But almost everything it suggested was either from ReadWriteWeb (naturally I have dugg a lot of RWW content in the past and it must have seen that), already promoted (like this -- which was also a couple of days old), or sitting high in upcoming.

Yes, it suggested things I like, but none of it was stuff I was unlikely to find. If I'm an avid Digg user, and if I'm using a tool like this you can probably assume I am, I probably have already seen everything high in upcoming and made popular in the categories I frequent. And any site that I have dugg very often (like RWW), I am likely a regular reader of. In fact some of the stuff it suggested was stuff I wrote (though of course there is no way it could have known that).

DiggFilter offered me a way to refine the results, but that just made things even worse -- all it did was strip out the non-RWW suggestions. DiggFilter's algorithm seems to put a lot of weight on stories from domains you've dugg often in the past, and it probably shouldn't.

DiggSuggest

DiggSuggest isn't nearly as pretty as DiggFilter, but functions in generally the same way: you put in your username and it gives you back suggestions. Along with suggestions it prints an enormous list of keywords culled from things you've previously dugg. The list is laid out like a tag cloud, but without text size or weight changes to show you which keywords appear more in your Digg history. That's unfortunate, because a real tag cloud would make the list somewhat useful, or at least interesting; as is, though, it is fairly useless.

The actual suggestions were very different from those given by DiggFilter. The most immediately noticeable difference is that most of the stories it suggests have very few diggs (under 10). Only a handful of its suggested material had made the main page.

DiggSuggest also recommended just one story from ReadWriteWeb (though it included that in my keyword cloud). The reason for both of these things might be that DiggSuggest only suggested links to me that were submitted in the past 5 hours. Because the links in their db are more recent than those that DiggFilter is working with, it probably has less RWW content, and less popular content to begin with.

As for the suggestions themselves, they were predictably tech heavy with a bit of politics sprinkled in. My guess is that DiggSuggest might take when you dugg something into account when looking at your Digg history. For example, a year ago when I wrote for a political news web site, I tended to digg more politics-related stories. Today, writing for a tech site, I tend to digg more technology-related stories. If DiggSuggest is as smart as I am giving it credit for, it may have looked at my history and decided I am more into tech now, but at one time I was into politics, so it might as well sprinkle in some politics recommendations as well. (Slight caveat: I do still digg political stories on occasion, just not as often.)

Conclusion

So which is better? For my money, DiggSuggest offered superior suggestions. While both gave pointed me at things I would be interested in reading, and DiggFilter may have even done so more consistently, DiggSuggest definitely was better at highlighting content I was less likely to see.

Of course, whenever it finally arrives Digg's official recommendation engine will likely be more sophisticated than either of these. It probably won't just look at what you've previously dugg, but also what people like you have dugg, what your friends have dugg, and what the people who have dugg stuff you've dugg have dugg. It could even also look at what else was submitted (or dugg) by people whose stories you've dugg in that past. I'm not sure that the Digg API offers outside developers enough information for a third party to come up with a recommendation engine that could be as sophisticated as Digg's may potentially be.

In the meantime, though, we have DiggSuggest and DiggFilter to help us find undiscovered content on Digg. Which do you like better? Let us know in the comments on this post.

How Well Does it Work?

Here's how Aggregate Knowledge describes the system for BusinessWeek.com:

"When a reader clicks on a breaking news story on the site, the Aggregate Knowledge Pique Discovery Window automatically provides user-driven content suggestions in the form of “More from BusinessWeek.” These suggestions are based on what visitors are actually reading across BusinessWeek.com."

I clicked some stories on the BusinessWeek.com homepage, and noticed a "More from BusinessWeek" list of links to the right of each story. However, none of these links seemed very relevant to the story. Check out this example from a story about Apple iTunes:

No Apple or even tech stories are linked to. Here's another example - about Russian police visiting BP offices. Curiously, this one lists an Apple story!

No Actual Content Analysis?

So based on my tests, it doesn't seem like there is much - if any - semantic analysis of the page content in order to come up with the "More from BusinessWeek" links. Reading between the lines of the AK quote above, this discovery system is based on clicks and not content.

It appears as if this is collaborative filtering - i.e. users who clicked X also clicked Y. This is basically the system that Amazon and Netflix use. For Aggregate Knowledge, collaborative filtering is still going to give interesting results. But how is it better than - for example - the 'Related Entries' plugin that we use here on ReadWriteWeb (which is based on tags, and so is much more closely aligned to the content itself). See bottom of this post for an example.

Surely for media sites a content discovery system that analyzes the content of a page, such as Reuters Open Calais does, would give better results. Please let us know your opinion in the comments.

It also positions MyStrands well as the leading force in recommendations technologies. While big guns like Amazon, Netflix and last.fm (owned by CBS) are all using recommendations, none has it as their primary focus. MyStrands describes itself, broadly, as a company that "develops technologies to better understand people's taste and help them discover things they like and didn't know about already."

In addition to the competition, MyStrands is also a leading force in a conference called the ACM Conference on Recommender Systems. MyStrands organized the first such Conference in 2006; and now it is organized under the umbrella of the ACM, last year in the US and this year in Europe.

As we noted in our previous coverage of MyStrands, we'd also love to see this company take a leading position in implementing open data standards. Although it seems to have lost the services of Scott Kveton, current Chair of the OpenID Foundation and formally MyStrands' Director of Open Platforms. Kveton is now at open identity start-up Vidoop, but back in December he was quoted as saying that MyStrands is "looking closely at APML, as well as working on some other 'open formats' for describing user taste data. The gist is, the users own this data and we want to give them as much control over it as possible."

Let's hope that the loss of Kveton hasn't deterred MyStrands from utilizing open standards in its goal to be the leading recommendations company.

Without further ado, here are our 10 picks, compiled from previous ReadWriteWeb posts:

MyStrands

How do you navigate a nearly infinite world of digital data to find the best content for your tastes and needs? Our collective answer to this question is in its infancy, but Oregon based recommendation service MyStrands has raised a whopping $55 million to build on the existing science of recommendation. Definitely the dark horse of the recommendation engines - one to watch.

MatchMine

MatchMine, a Massachusetts company building a cross-platform media recommendation engine, received a $10 million investment from The Kraft Group. The company released an early product called MyMovieMatch in July. See RIA expert Ryan Stewart's review of the original product for background from this summer.

Zync

Zync, a Massachusetts-based startup, operates a local event recommendation engine based around the city of Boston. The site currently lists 30,000 events across 20,000 venues. And even though it only has 355 users, they have amassed almost 9500 ratings.

According to Zync, their recommendation technology uses patent-pending algorithms to recommend events, activities, and restaurants to users based on the input of other, like-minded people. Theoretically, as with any peer recommendation system, this one would get better and more accurate the more people use it.

SeeqPod

The team behind SeeqPod, a music search and recommendation engine, believes strongly in what they call "playable search." SeeqPod trawls the web, indexing all the music files it finds, and then offers them for playback direct from that location. The company knows that because they are not hosting any music files, but are merely offering links to them, they can neatly sidestep copyright and legal concerns.

Scouta

Scouta is a web app that provides you with media recommendations, based on preferences and interests you display by your selections within the application. If that sounds complicated, think Pandora, but for all media on the web (including media available outside the US). Or think Last.FM without the fuss about neighbors. To be honest, neither of those comparisons is quite right either. It's more like YouTube, except all the side column content is actually interesting to you.

TuneExplorer

Music recommendation and discovery engines are hot stuff but what if you could use some of the same juju to better organize the music you already have in your collection? The newly launched Veenix TuneExplorer for Mac does just that. By looking at qualities the company says include "pitch values, pitch variance, fundamental strengths, and a host of other sonic qualities" - the program acts like Pandora within your music collection.

TheFilter

The Filter, a social music recommendation service backed by rock star Peter Gabriel, has released a new version of their software - featuring an improved user interface, a Facebook app and a partnership with Nokia. The Filter is a "playlist creation suite" for iTunes, iPod, iPhone and Apple TV. It works across Windows and Macintosh and it basically allows you to build playlists from the music stored on your PC, Mac, iPod or Nokia mobile phone.

The Filter's user base is reported to be growing at 25,000 a month. The engine can identify 5 million songs, 4.5m of which have clips (short samples). The Filter works by using Bayesian mathematics and it was developed by physicist Martin Hopkins.

Criticker

Born out of a closet dislike for "Shrek 2," Criticker is a new movie review community and recommendation engine that aims to match users with like-minded individuals who share the same cinematic taste. Once you've rated 10 movies at Criticker it begins to form what they call a Taste Compatibility Index (TCI) that matches you up with not only other users, but also professional reviewers who share your taste in movies (though, we found that site really doesn't start delivering usable results until you've rated around 50 flicks).

FeedEachOther

FeedEachother is an RSS Reader built by a former developer from Yahoo! Answers and another now at craft social network Etsy. The interface will feel very familiar to anyone who uses Facebook or Google Reader. The service does a good job of communicating for novice users while offering a feature set that power users will really like.

FeedEachOther recommends feeds "similar" to the ones you're subscribed to. Recommendation engines are a key way to leverage the network effect of distributed nodes of knowledge - ala social apps online. Big value there for discovery of high value information sources.

StumbleUpon

StumbleUpon is a "personalized content discovery" service, which has grown very popular on the Web. Its main feature is serendipity, finding new webpages by clicking through from other pages 'stumbled' by users. The app is now owned by eBay and it's unknown what they might do with StumbleUpon, but recommending new items to buy might be on the cards.

Conclusion

Now it's your turn: recommend some more recommendation engines in the comments.

Recommendation Resources:
*The Art, Science and Business of Recommendation Engines
*RWW Recommendation Industry Feed Favorites OPML file (save link)
*RWW Recommendation Industry Feeds - Best of Feed (copy and paste to your reader)
*Click to preview the above feeds before subscribing (pop-up window)
*RWW Open Data Sites Search (Visit and Bookmark)