Glue's synaptic web-esque technology is based on a user's browsing across common sites such as Amazon, Wikipedia and YouTube, and those visits and any interactions (comments, "likes," etc.) feeding back to automatically create a taste profile and a web of affinity with other users and recommendations of other items or content across about a dozen categories, including music, books and movies. So, can this be done without violating users' privacy or - worse yet - frustrating and boring them into attrition?

AdaptiveBlue VP Fraser Kelton thinks both aims can be accomplished. In a phone interview this morning, he shared that Glue has a three-tiered set of privacy controls to ensure that items are shared only when a user wants them to be. As the privacy/permissions inverse of Facebook Beacon, an ill-fated system for socially powered product recommendation, the Glue system is completely opt-in for users and opt-out for publishers.

Glue's all-new user profiles are rolling out today at GetGlue.com.

"Nobody's confortable sharing all of their stream," said Kelton. "Financial sites, adult content - we don't want any of that stuff shared." In addition to having a set list of sites from which to pull browsing and interaction data, Kelton said, "The secondary privacy controls can prevent visits to a website from being shared unless the user has a specific interaction on that site. You can also opt to have a private profile so only your friends can see the items you interact with." Finally, Kelton told us that users can delete or edit their own shared content on an item-by-item basis.

Our other concern was whether Glue would be a semantic web geek-fest or whether the average end user would also be able to use, enjoy, and return to the site. Kelton said the site went over well with non-techie beta testers, many of whom started out using the GetGlue.com website and eventually also installed the add-on for an even richer experience.

One of the factors we think makes Glue so much fun for end users, from geeks to complete Luddites, is its Foursquare-like system of achievement and rewards. "We're big fans of game interaction driving the user experience," said Kelton, who also revealed that the product team considered Xbox 360 as inspiration for the Glue system. Users can earn stickers on their profiles based on how many items they browse or interact with, how many users discover content through them, and how much of an expert they are on a particular topic.

Glue's game-like system of badges or stickers as rewards for user interactions and discovery makes for addictive browsing.

Each item indexed by Glue can have a guru, and only one guru at a time. The result is friendly competition, increased interactions and a very sticky website that's compelling and entertaining for even the least technical end user.

The site also allows for content to be pushed to a user's existing social networks, such as Facebook and Twitter. Here's a demo video showing more about how the site works:

As previously mentioned, the user experience of Glue - both the site and the revamped add-on - is one of synapses and nodes, not back-and-forth browsing and search for content. As data and content from all around the web is associated with films, books, music, and other items, users are able to quickly scan and consume related content without further navigation. This type of curated experience is something the AdaptiveBlue team calls "shortcut search."

As a social network, GetGlue.com is a hub; but the way it curates content makes it a series of spokes, as well.

"You don't have to go to Rotten Tomatoes or YouTube. We know there's a subset of information users want, and we're bringing it in," said Kelton, who is very excited about Glue's position on distributed social networking and the semantic web. "Jeremiah [Owyang] uses the term 'social colonization.' I also think we're implementing what Charlene Li talks about when she uses the phrase, 'The network is all around us.'"

Alex Iskold (@alexiskold), the founder of AdaptiveBlue and occasional RWW feature writer, was nice enough to sit down with me and give me a test drive of the new version of Glue. We started out covering the basics (which we wrote about here) and moved on to the new features:

Connected Conversations

Building on the concept of being able to share thoughts and opinions on things with your friends on Glue, regardless of the site those things are found on, is taken to the logical next step with the addition of conversations. Now, if you see that someone has commented on something that you are looking at, or have an opinion on, you can add a comment to their opinion. In turn they can comment back, or others can join in on the conversation. Through these interactions, you will be exposed to new people who perhaps came to the conversation from a completely different web site, Wikipedia for instance, instead of Amazon, but are using Glue to transcend the social boundaries of these sites.

Smart Recommendations

Being a contextual network that uses semantic technology to gather information and trends, Glue now aggregates this data and can present what books, movies and music your friends like the most instantly. Creating this recommendation data is done automatically as people use the Glue application by indicating what they like. The lesson here is, the more you use glue, the better a resource you become to your friends who also use the service.

Web Wide Top Lists

Expanding the snapshot to encompass the entire Glue network, web wide top lists contain the aggregate weekly activity of all Glue users, pointing out larger trends and new music, movies and other things that are just being discovered. This is also a great place to find new Glue users to add to your network, and find active conversations on comments that have been made by the Glue community.

Questions and Answers

At the end of the presentation, I was able to ask Alex some questions about Glue. My biggest concern was one of privacy; How could I be assured Glue didn't watch everything I did in my browser, looking for the next semantic morsel to snatch up? It turns out the Glue addon is only active on certain sites, and does not communicate with the server until it finds a music track, movie or book that can be enhanced with Glue. Put simply, if the Glue bar doesn't appear on a web page, the Glue plugin is inactive. Alex also pointed to this web page that lists the sites that do trigger Glue.

Final Thought

People who have been using Glue since it came out will tell you it is addictive. Now, especially with the ability to have conversations around the books, music, movies and other things that you enjoy, we are sure there's no way to not get addicted and quickly find like-minded folks who can give you great recommendations on things that you will love.

One of these evolutions started quietly in 2008. We are witnessing the rise of a new kind of web: contextual. You might not have heard or thought about it much yet, but you are already using it today. Search remains the killer app on the web, but context is quickly become a viable contender. Why? Because context is what happens instead of search.

Until recently on the web, most sites have not been software - only data, a bunch of flat HTML pages. The software that you used to look at them was, of course, the web browser. The problem was that the browser had no idea what the pages contained, and it did not know what you were doing. Because the browser could not infer your context, it could not help you explore related and relevant information.

The contextual web experience is fundamentally different because there is an understanding of what the user is doing. The combination of the information on the page and the user's behavior creates the context. Once you understand the user's context, you can be more helpful. So, contextual technologies have the potential to bite into the pie that today belongs to search, because it is able to bypass search.

Consider the difference between looking at the IBM home page and a movie page on Netflix. The context is very different, and so likely is the user's intention. The user may be looking at IBM because she is looking for a job or researching IBM products. When the user is looking up a movie, it is because she is thinking about renting it.

The Key Properties of The Contextual Web

The Contextual web will happen when browsers and websites evolve to recognize what users are trying to do. It is the web with less choice and more meaning, where instead of Googling all the time, we Google once and then the rest of the information is available to us automatically, based on our current context.

Here are the key properties of the contextual web experience:

• Relevancy: understanding the user's context better drives content relevancy.
• Shortcuts: contextual shortcuts reduce the need for raw search.
• Personalization: context is based on user intentions and history.
• Remixing: relevant information from around the web is instantly available.

What are some specific examples of contextual technologies that are improving our web experience right now?

Markup Technologies

One of the keys to inferring user context is understanding the underlying information that the user is looking at. This is why the contextual web is related to -- and, to be more precise, is powered by -- semantic web. We have written a lot here on ReadWriteWeb about semantic technologies. Notably, we discussed the difference between the top-down and bottom-up approaches to semantic web, both of which are important for understanding the contextual web.

The bottom-up approach to context is about annotating pages. For example, all modern browsers can detect if the page you are looking at contains an RSS feed. This happens because the browser looks for a tag in the head of the page, which declares type . The fact that the page offers an RSS feed creates an obvious context: subscription. So, the browser then invites you to subscribe using your favorite RSS reader.

Another form of markup that has been discussed recently is microformats, which offer an XHTML-compliant way of embedding metadata about people, places, events, and reviews in existing web pages. Even though microformats are not ubiquitous today, there are clear benefits to using them. The image below is from an excellent post by Mozilla UI lead Alex Faaborg about leveraging microformats in the browser.

Building on the hAtom microformat are Web Slices, introduced by Microsoft in Internet Explorer 8. Web Slices enable publishers to notify users when the information in their web pages changes. For example, Weather.com can create a Web Slice that tells the user when a local weather update is available. eBay can deliver a Web Slice that notifies the user when the price of an auction changes. Similar in concept to RSS, Web Slices focus on updates in part of the web page, enabling publishers and users to communicate directly via the browser.

There are other markup formats that help provide context. For example, popular add-on Cooliris offers a markup format for signaling that a site contains images. By placing a bit of XML code in their home directory, site owners enable users to experience their images using the stunning 3D visualization developed by Cooliris. Another markup format, developed by AdaptiveBlue [disclosure: this is the company I founded], is called ABMeta. This format allows publishers to annotate pages that contain information about books, music, movies, wine, restaurants, stocks, and other everyday things.

All of these markup-based approaches face the same issue: publishers have to do the work of actually annotating the pages. And while the semantic web community has been very vocal about the benefits of annotation, the majority of the web is still flat HTML.

Page 2: Widgets

Widgets

In the meantime, we are seeing a surge in applications that deploy a top-down approach to inferring user context and being helpful. Instead of relying on markup in pages, these applications use heuristics and APIs to recognize the information that the user is interacting with. Based on their recognition, these smart tools then offer contextual shortcuts to the relevant bits of information. Speaking broadly, today there are two primary categories of top-down contextual applications: blog plugins and browser add-ons.

Blog plugins offering a contextual experience have been around for some time, starting with preview technologies. One of the first recent unsuccessful attempts at a preview technology was delivered by Browster. Despite its failure, its successors, CoolPreviews, SnapShots, and Apture, have done much better. While CoolPreviews literally focuses on a preview of the page, SnapShots and Apture deliver a compact summary of the information behind the link. The basic premise behind the previews is sound: if the user is not interested in what is behind the link, the preview can save the user an unnecessary click. Because the user context (in this case, the link) is known, relevant information can be brought to the surface and delivered to the user instantly.

Closely related to previews are widgets that offer contextual shortcuts. Like previews, these widgets work by being anchored to links. Examples of this technology include Yahoo! Shortcuts and SmartLinks from my company AdaptiveBlue. Instead of providing a preview of the underlying content, these technologies offer links to related content around the web. The example below shows a SmartLink on a New York Times movie page, which invites the user, for example, to buy the movie on Amazon or rent it on Netflix.

Our last example of contextual widget technology comes from Colorado-based Lijit. Lijit has developed a search technology that focuses on the individual. You claim your presence around the web, on blogs, Twitter, Flickr, etc., and then others can search your pages on specific topics. Say you are on someone's blog and want to do a search on that person; it makes the most sense to search through this person's pages first; after all, you are in this person's context.

Remarkably, Lijit offers another contextual nugget, called Re-search. It is simple, but a great example of the power of context. If you search for a term on Google and end up on a blog that has a Lijit widget, a header appears with additional results and a prompt to search for more. Lijit automatically infers your context, does a search, and offers additional helpful shortcuts.

Page 3: Browser Add-Ons and Conclusion

Browser Add-Ons

Widgets and markup technologies are making inroads into the contextual web, but an even bigger inroad is being made by browser add-ons. Since Firefox introduced its platform for delivering additional features to the browser, many thousands of browser extensions have been developed. A lot of these extension focus on enhancing the browsing experience by leveraging context. Probably the granddaddy of them all is Greasemonkey, an extension that allows users to install bits of JavaScript that alter the look and content of web pages they visit. We wrote an introduction to this popular add-on and recently followed up with a post on 7 New Greasemonkey Tweaks.

While Greasemonkey scripts are mainly regarded as experimental tools for power users, quite a few other extensions are aimed at building a business around contextual enhancements to the browser. One example is a company called WebMynd, which focuses on enhancing Google search results. WebMynd's extension automatically pushes down Google ads (clever!) to insert its contexual gadget. It allows users to simultaneously search posts on Twitter, book matches on Amazon, video matches on YouTube, and many other sources. An example of what shows up when searching for "semantic web" using the gadget is shown below.

We have mentioned Cooliris several times already. The most popular Cooliris product is the browser add-on that shows stunning 3D views of photos and videos around the web. Cooliris works contextually by automatically recognizing popular photo sites like Flickr, Google Images and Photobucket. Another contextual add-on that works by recognizing content on pages is Glue. This add-on shows you friends and other users who have visited the same book, music, movie, restaurant, wine, and other pages around the web. The context here crosses the content you are viewing with your social graph. Unlike lifestreaming applications that deliver you information about your friends out of context, Glue only brings to the surface information that is relevant to your current context.

A very different example of a contextual web add-on comes from Slovenia-based Zemanta. Its add-on makes it easy for publishers to add contextually relevant links, photos, and video to their posts. Zemanta works by applying its semantic engine to the body of the post and automatically recommending related content. Using Zemanta, bloggers are able to instantly add relevant content to their blog posts, in turn creating a relevant web experience for their users.

Our last two examples of contextual add-ons are Zentact and App Discover, both recently launched. Zentact -- developed by the folks who brought us another contextual phenomenon, MyBlogLog -- aims to solve a problem of staying in touch with people who matter to you. First, you import and tag your email contacts with different tags that reflect their interests. Then, as you browse, Zentact pops up a reminder to contact a person if it deems the page relevant based on the tags. It is a simple yet powerful example of the contextual web.

App Discover works by recommending related applications to the sites that you are visiting. Today, it requires publishers to add markup to their sites, but one can easily imagine how this technology could work top-down. The example below shows App Discover recommending TweetDeck when the user navigates to Twitter.

Browsers

Without a doubt, web browsers are in the best position to deliver the user's contextual web experience. While add-ons are taken in a variety of directions, the two dominant browsers, Internet Explorer and Firefox, have already incorporated the basic contextual experience: shortcuts. Internet Explorer 8 features technology called Accelerators.

According to Microsoft, Accelerators give you ready access to the online services you use everyday, from any page you visit. Accelerators are defined as little chunks of XML based on pre-defined variables and defined by the browser. Some examples of variables available to Accelerator are the active URL, the active domain, and selected text. The most common action that Accelerators do is perform contextual search based on the user's selection. Another common Accelerator function is the in-place lookup, such as looking up a map based on a given address.

The problem with Accelerators is that they are not really based on a selection; that is, they lack semantics. When you highlight an address, you still have to decide which Accelerator to use. If you have dozens of Accelerators installed, this quickly becomes hard to manage. Firefox has recognized the issue with the menu-driven approach and instead offers its contextual technology via text. Called Ubiquity, this contextual technology is still only available as an add-on today but is likely to be a part of Firefox core soon.

Marketed as user-generated mashups, Ubiquity is actually a contextual technology based on language. As with Accelerators, the user is able to select a piece of text and then invoke Ubiquity and type a command. In the screenshot above, Ubiquity is used to insert a map into an email. Hundreds of Ubiquity commands have been implemented to date. ReadWriteWeb has already written about Ubiquity commands.

Will Context be the Future of the Web?

So where is all of this heading? Are these technologies signal or noise? We believe we are witnessing the birth of a fundamentally new kind of web, a smarter one, a contextual one. Unlike the old web we are used to, this one understands what we are doing and helps us. It is a web in which we search less and find relevant content faster. This new contextual web is still very young and unevenly distributed, but it is definitely here.

The fact that these contextual technologies are springing up is not accidental. The Contextual web is made possible by our push into semantic web and the rise of web services/API culture. The combination of basic semantics and API is fueling all of these contextual applications. Bit by bit, the web is getting smarter, friendlier, and more enjoyable.

Web browsers are in the best position to deliver these new contextual experiences to users because of their wide reach. The fact that Microsoft made Accelerators its flagship feature for Internet Explorer 8 and that Mozilla is putting much effort into Ubiquity tells us that contexual browsing is a priority. This is really good news, because these technologies bring great benefits to the user.

It is great to see that in these tough economic times, evolution is brewing. The tremendous effort that all of these companies are making with contextual technologies is beginning to come to fruition. It may be that we are seeing glimpses of what the next generation of the web will be like.

And now, as usual, let's turn the tables. Tell us examples of your favorite contextual browsing technologies. Do you think context will be play significant part in the next evolution of the web?

"When you type in a search today, you get a list of URLs, and they are not very informative," said Rajeev Rastogi, vice president of Yahoo Labs Bangalore, where the technology is being developed. The lab is working on automated information extraction which goes into URLs and extracts relevant information. For a hotel, the additional information returned may include an address, phone number, map to the hotel, and its rating. For products, you may get an image of the product, the name of the manufacturer, and the price.

At first, that sounds a lot like Google's Universal Search which returns results from across the Google Search verticals (images, news, etc.) when performing a search at google.com. The difference is that the abstracted information will appear under the URLs listed in a fashion that's very similar to Yahoo's SearchMonkey experiment, a technology that allows independent developers to enhance their site's appearance in the Yahoo search results. This is done by using semantic markup (microformats, RDF), standardized XML feeds, APIs (OpenSearch or other web services), and page extraction.

As great as SearchMonkey is (we included it in our top 10 semantic web products for 2008), not everyone is using it. Says Rastogi, "Clearly we don't expect that everybody will adopt SearchMonkey, so this 'rich results' piece is our in-house effort to automate the information extraction for large classes of web sites." In other words, Yahoo will "SearchMonkey up" the web sites for you.

New Technology Will Also Recognize Intent

In addition to changing the appearance of the results themselves, the new technology will also offer users help with refining their queries. How closely this aspect ot the technology will resemble Yahoo Glue is yet to be revealed, but it could have some similarities. What is clear, however, is that the feature will go further than current technologies like Yahoo's Search Assist which auto-completes search queries as you type. Instead, the new technology will prompt users to narrow down their queries by recognizing user's intent. In order to determine what that intent is, it will examine previous behavior like the user's prior web searches, visits to various Yahoo web properties, and "other information."

Although that mysterious "other information" cited in the PC World article makes us curious about privacy issues and tracking cookies, the idea is intriguing if you can get past the unsettling feeling that your search engine will get to know you a bit too well. We imagine this to be like an even smarter version of Google's Auto-Correct feature, something that currently gives you the option to search on the correct spelling of a word instead of the misspelling you had typed in.

So, unlike Google, which simply asks you "Did you mean: porshe" (when you queried on the word "porsh"), Yahoo's technology may, in theory, ask "Did you mean: porshe 911 turbo in blue?" That's just freaky.

In regards to how MBA students are grokking Web 2.0, Bud made this observation:

That doesn't surprise me and I think you'll find the same ratio of Techies are deeply interested in business administration! :-) But it's all about bringing Web techies and business folk together, so that each has an appreciation of the other - and we can build partnerships. For MBA and business people, there are opportunities to launch businesses in the Web 2.0 'space'. Likewise for Web techies, we need people with business nous to get our ideas and products off the ground.

So I think Bud's program is a fantastic way to introduce business people to Web 2.0 technologies and I'll be watching with interest to see how it turns out.

Poweryogi (real name please?) also replied to my post - he's the MBA student who is thinking about launching a Web 2.0 venture. In a post today, he said he's a bit pessimistic about how MBAs can "make an impact" on Web 2.0. He wrote about a conversation he had with two other MBA students - "one a McKinsey-bound consultant, and the other a UBS-bound banker." When poweryogi steered the conversation towards Web2.0 and blogs, he noted that the banker "seemed disinterested" and the consultant wanted to know how blogs differed from IM.

I thought poweryogi's last comment was the most revealing: he noted that MBA students have "little time to experiment". That's entirely justified and actually I see no reason why MBA students should experiment. There has to be a business reason behind most things they do, so it's our job as technologists and Web people to show them where the business value is.

Bud Gibson posted a similar comment on poweryogi's blog: "...when you are on the cutting edge, you have a lot of selling to do". Exactly and in fact that's part of my job(s). Just recently I was complimented by someone in my day job on my "translation skills" - meaning from Geek to Business and vice versa! That's kind of what I do here on Read/Write Web too.

Poweryogi is one of those 1-5% MBA people I think, so hopefully he continues to explore Web 2.0 and helps explain the business value to his classmates. And who knows, maybe one day I'll be approaching him to lend me some venture capital ;-)

If had to use one word to explain how I see the relationship between Technologists and Businesspeople, it's symbiosis. We're two different species and each group has its own specialities. But together we can develop a relationship of mutual benefit.

ahhhh, what a nice way to end a post - with a biological metaphor!

This week: business folk getting interested in Web 2.0, Adam Curry podcasting from 2.0 perspective, cool Web 2.0 'mini-apps', wrap-up of the adverts in RSS debate, Bosworth's Web of Data.

From MBA to Master of Web 2.0?

I get accused of being too geeky sometimes on Read/Write Web (no argument there!). So I'm on the hunt for more business-related Web 2.0 stories. I do believe that Web 2.0 is starting to permeate into mainstream business culture - perhaps from the bottom up, i.e. from business schools. For example this MBA student is looking "to mesh the classroom teaching of proven theories with the rapidly evolving wild west of what's being referred to as Web 2.0." He or she (no real name provided, so I couldn't tell) goes on to say:

That's encouraging to hear and I wish that MBA student the best of luck.

Of course the other way business folk can be introduced to Web 2.0 is to use the tools. For example Boris Mann recently talked about weblogs and wikis to a local MBA class.

Curry Podcasting: Implications for Web as Platform 

News this week that Sirius Satellite Radio Inc. is launching a podcasting show, to be hosted by ex-MTV star Adam Curry. It'll be a four-hour weekday show, featuring a selection of amateur podcasts handpicked by Curry.

Although it could be argued this is more about broadcasting than podcasting, given Curry's background with MTV, I think this has huge implications for Web 2.0. It's a chance for enthusiastic podcasters to get their material heard by a much wider audience. Indeed it's not far-fetched to suggest that a small percentage of talented podcasters will bootstrap their way to their own professional radio/podcasting shows - perhaps even becoming stars.

I'm sure Curry will unearth some real podcasting gems, over time. Although I suggested he may be the Casey Kasem of Podcasting in my ionrss.com piece, it's probably fairer to say he could be the John Peel of Podcasting. Peel was a music-lover who discovered and recorded some amazing new bands, who probably wouldn't have made it but for Peel. There's an opportunity for Curry to do the same for Podcasters all over the Web.

Cool Web 2.0 Mini Apps and Services

I'm doing my best to avoid writing about The Big 3 Internet companies this week, so to extend that theme I thought I'd list out some neat new Web 2.0 things developed by individuals or small companies. Here are some I discovered this week:

- RSS Mix: a feed remixer
- Foundcity: "allows everyone in a city to map the interesting things they discover throughout the day to a dynamic online map"
- Adactio: uses APIs to collect scattered pieces of Web content into one place (see my review here)
- Backpack: amongst other things, transforms emails into functional web pages. Got a lot of blog buzz this week, including from me.
- airWRX: a content-creation workspace that runs from a USB flash drive
- hReview: an open standard for reviews (see Phil Pearson's implementation at the NZ Coffee Review site).

Feel free to email me (see my site's menu) if you have a new Web 2.0 app or service you want me to take a look at. 

Ads in RSS Round-Up

I've been covering this issue on ION RSS. In terms of the Web as platform, I concluded that RSS is essentially equal to HTML as a publishing format. That is, RSS is a first-class citizen of Web publishing. People can and will put anything they want into an RSS feed, just as they do with webpages.

But, as always, each 'end-user' will decide for him or herself whether ads in feeds are acceptable. It's easy to unsubscribe from feeds and that's part of the beauty of Web 2.0 - users have control over their Web experience.

Techie Time: Bosworth's Web of Data

Adam Bosworth recently gave a speech to the MySQL Users Conference 2005. Bosworth is a former Microsoft web wizard, but nowadays he casts his RESTian spells as a Google employee. He's known for his evangelism of simplicity and 'sloppiness' in designing for the Web. For example, here's his view of RSS:

For a design and business-oriented take on the 'Web of Data' theme, check out Web 2.0 for Designers, the Digital Web Magazine article that Joshua Porter and I wrote. If you're geekily inclined, I also recommend you check out Bill de hÓra's and Dare Obasanjo's posts in response to Bosworth's speech. Also David Megginson has a very techie post on this theme.

Summary

That's a wrap for another week! Hey, I should make that my catch phrase... I hope you're still enjoying these posts and as always, I value any feedback. Feel free to email or leave a comment.