Search 2.0

Google has taken a big step into the social search arena. Google now allows you to create customized search engines. A customized search engines is a Google search that is restricted to a specific list of web sites that you specify, like search 2.0 company Rollyo’s searchrolls. An interesting feature is the ability to invite others to help edit and customize in a cooperative manner with you, and to have volunteers offer their help too. Currently you only get 100 invitations, but the number of people that can volunteer is unlimited. Google manages the entire invitation and volunteer process for you, and you do have editor privileges over anything submitted to your customized search engine.

You can also refine your target web site list by adding keywords to highlight certain topics a particular web site offers. This allows you to emphasize certain keywords that are hard for the search engine to pick out. For example, suppose a particular web site is great for reviews. In an normal search, the word review is used in many different contexts leading to search results that have a lot of web pages that don’t actually contain reviews. By refining a web site in your search list with the word review, you are telling Google that it is specifically a review site.

Google has added a powerful motivation for people to use this new program. AdSense. Once you have created a customized search engine, you can put a search box on your web site and make money from it, via the AdSense ads placed along side the search results. It will be interesting to see if the other social search engine companies like Wink, Eurekster, and RollYo follow suit and allow contributors to monetize their search boxes. I would think they will have to in order to survive.

A blog post by Matt Cutts, where he reviews Google Co-Op, is the seed for today’s post on Google & Social Search. Google Co-op is a volunteer program where human editors contribute links to help display relevant search results to the searcher. The volunteers do this by tagging topics with keywords that tell Google when to display contributed links in response to search queries, queries that use one or more of the tagged keywords. These keyword tag files are currently in XML. You can learn more about them and see a sample in the Google Co-op Topics Developer Guide. Google Account users subscribe to different providers from the Directory page. As long as the user is logged into their Google Account, the next time they make a search query Google will:

• See what providers the user has subscribed to
• Scan them for links that have tags relevant to the user’s query
• Add those links to the top the search results displayed to the user

At the time I wrote this post there were 13 providers in the Google Co-op Directory.

A new feature recently rolled out is the refine your results feature. When you search Google, depending on the topic, you may see links at the top of the search results to other categories related to your search. If you are subscribed to one or more Google Co-op subscriptions, some of the links will come from that source if there are applicable topics. This feature can guide you to a sub-topic or related topic that satisfies your search.

Google’s almost stealthy way of rolling out new features is fascinating to me. Here’s a brand new feature that will have a major impact on how Google’s search results are affected by human editors, and I would wager 99.99% of Google searchers have no idea it exists; yet if they did they could choose to benefit from it today. The human editing and annotation side overlaps some of the purposes of an entity like Wikipedia; not on the content creation side but in how editorial opinion can affect Google’s search results. (If you don’t think Wikipedia is affecting Google’s search results, albeit indirectly, you haven’t really done any searching lately.) The “refine your results” feature, although not algorithm based clustering like that used by the Clusty search engine (and others), does provide similar functionality to the user. Finally, human edited directories like DMOZ (ODP) and others will be affected if Google Co-op links are weighted more heavily than the results from authority directories.

Here’s a giant Search 2.0 event that has managed not to draw huge attention from the web community (yet). The following is speculation, but in my opinion Google Co-op also sheds some light on the motivation behind Google’s new editorial opinion patent, which discusses the use of human edited query themes in conjunction with improving search results. Even more interesting is that they have made available a subscribed links API, which allows subscription content providers to add content programmatically to the Co-op.

Experienced search engine fans will quickly wonder how they intend to control spam in the Google Co-op; especially since they are providing an API. Here are some of the anti-spam related references I found in the Co-op online documentation:

“Contributors listed in the directory are people and organizations the community has found particularly useful and of high quality. The Co-op Directory does not include all Co-op contributors.“

“As with the style guidelines for Responses, if your results flagrantly violate the principle of triggering only on relevant queries, Google reserves the right to disable their display entirely, as per the Terms of Service.“

I anticipate that Google will need more aggressive anti-spam policies and mechanisms such as user flagging and probationary periods to prevent Co-op spam in the future.

Keywords: google, google co-op, matt cutts, google directory, dmoz, odp, search 2.0, search engines

Technorati Tags: google, google co-op, matt cutts, google directory, dmoz, odp, search 2.0, search engines

Recently, GoDaddy CEO Bob Parsons made a revealing post on his blog on why he chose to pull GoDaddy’s public offering (IPO) off the stock market. GoDaddy is a leading domain name service provider and is one of the better known Internet companies that has managed to survive. His comments on the current IPO market lead me to wonder if, at least in the short term, investment funds for new Internet startups will begin to dry up. Although I still see reports of Search 2.0 companies successfully raising millions of dollars in investment capital, his comments have a somewhat chilling effect. In his blog post, as part of his explanation for pulling the GoDaddy IPO, he offers these sobering words on the current state of the IPO market:

“We have war and escalating hostilities throughout the Middle East, with no end in sight. Oil prices are skyrocketing. Tech stocks, in particular, are once again taking a beating on Wall Street, due in part to some investment banks cutting their ratings on the U.S. technology sector. Rising interest rates have played a key factor. Their steady rise over recent months has put adverse pressure on stocks overall.“

I sincerely hope this does not affect the current and new Search 2.0 startups seeking new capital. Many of the new offerings have great potential and it would be a shame if such powerful search improvement ideas such as collaborative search (wisdom of crowds), user personalization, community policed spam filtering, never saw the light of day because the money wasn’t there to explore them. For a more detailed read on the current IPO market, read the article by USA Today, “Investor Pain Overload“, which Bob points out in his post.

Keywords: search 2.0, search engines, godaddy, ipo, investor, stock market, collaborative search, wisom of crowds

Technorati Tags: search 2.0, search engines, godaddy, ipo, investor, stock market, collaborative search, wisom of crowds

Ebrahim Ezzy, the lead developer and co-founder of Qelix Technologies, the company that makes search 2.0 contender Qube has an interesting round-up post on Read/Write Web. In it he highlights several of the leading Search 2.0 companies including:

• RollYo - a social search engine that allows you to dynamically create lists of web sites to restrict your searches too. These searchrolls can be shared with others to help propagate the spread of knowledge embedded in the searchroll.
• Wink - collaborative search service that uses the user community to police spam and tag search results.
• Lexxe - encourages you to enter your search query as a question and uses natural language techniques to provide you with the most relevant matches; works best with short queries but definitely does a good job.
• Clusty - search engine based on Vivisimo’s clustering technology that displays search results in folders that are centered around different but related topics.
• Eurekster - the creator of swicki’s, small information kiosks hand crafted by users that serve up information on tightly focused topics.

My favorite part of the post is his third criteria for applying the Search 2.0 label to a company. It’s the clearest, most elegant definition I’ve seen yet:

“Third-generation search technologies are designed to combine the scalability of existing internet search engines with new and improved relevancy models; they bring into the equation user preferences, collaboration, collective intelligence, a rich user experience, and many other specialized capabilities that make information more productive.”

His company’s flagship product Qube is a desktop application that provides advanced Internet search capabilities. I have not looked at it yet but here is what it does in the company’s own words:

“Qube v2 centers around the concept of browserless search, blending third generation websearch capabilities with user collaboration, progressive advertising & social networking.”

Take a look at the post, it’s a good read. If anyone has experience with Qube, please leave a comment or send me an e-mail.

Keywords: search 2.0, search engines, search, google, yahoo, msn, qube, clusty, wink, lexxe, eurekster, rollyo

Technorati Tags: search 2.0, search engines, search, google, yahoo, msn, qube, clusty, wink, lexxe, eurekster, rollyo