Semantic Search Engine Hakia Now Says It Can Filter Results By How Credible They Are
by Erick Schonfeld on October 6, 2008

On the Internet, nobody knows your site is a dog (to paraphrase the famous New Yorker cartoon). At least not yet. Semantic search engine Hakia wants to change that. Ask.com is not the only search engine rolling out a redesign today. So is Hakia, which is introducing tabs to its search interface. One of the tabs is “credible sites.” These are results from sites that have been vetted by librarians and information specialists (although anyone can suggest sites). So far, Hakia has built out a directory of credible sites around health, medical, and environmental issues.

The “credible” results tend to come from government, university, medical, and news sites. For instance, here are the credible results for “green buildings” and “common cold.”

The idea is if your site is a dog, it won’t get on the list. While this white-list approach could improve the quality of results, it also seems way too easy to game. Any spammer can try to get their site on the truthful and authoritative list. And they will.

Hakia’s redesign includes some other new tabs as well. Namely “news,” “images,” and “meet others.” (No, this third one is not a semantic dating service. It is a social feature that lets you join groups and discussion forums around the same topics that you are searching for). The results are all based on semantic matches within each domain. (See the screenshots below). But I am not sure how much better they are than Google’s. For instance, here is Hakia’s image search results for “global warming”. Now try the same search on Google image search. In both cases, the results are underwhelming.

Advertisement

Comments rss icon

  • Homepage does not have a title.. LOL

    And one random and first search for “What is a blog” term for credible sites, gave one result and actually, that too was irrelevant.
    Ask.com is a lot better then!

  • I think this is the way to go – the future of search and news will revolve around quality and credibility, as opposed to the popularity driven model we see today. Of course, credibility has subjective and objective components, and there are challenges to empirically measure “quality.” But as one of the startups also focusing on credibility driven results (for us, it’s in the news space), I wish Hakia the best of luck!

  • Junk. Just queried “How to use a blog for lead generation” and got recommendations for treating lactic acid, librarian trends, and other utterly off-topic stuff. Worse, the pages it linked to gave me 404s.

  • A couple of the queries I tried gave me problems like Tester and Chetan had, really anything with long phrases or complete sentences. But in some cases it was helpful. For instance when I searched a couple of my favorite TV shows it categorized things by official sites, news, what channel its on, cast and crew, etc.
    I don’t know that they are ready to promote themselves with all the bugs needing to be worked out, but they certainly have a good start on this service.

  • This should be the future of search. Weeding out the crap and leaving us with better overall searches. Thanks.

    Craig
    http://www.budgetpulse.com

  • Dear all,

    Thank you for giving hakia.com a try! We are currently adding content- general Websites and Credible Sites- and the performance may be uneven. For a glimpse of our capabilities, please try us with your health and medical queries.

    Our goal is to offer an alternative to popularity-based systems. Google is one of the pioneers of Web search and turned 10 last month. It will take us time to expand our coverage for a fair comparison to that of Google. That said, we think this is the direction Web search is going: offering semantic relevance and guidance about credible Web content.

    Again, thanks for the feedback!

    • Farrah,

      Good luck with the launch today. I’ll be seeing you at the Web 3.0 Conference & Forum – let’s grab a drink and talk for a few minutes between sessions – I’d like to introduce myself and learn more about Hakia.

      -Allan

  • For a health search engine, it is critical to be able to weed out spam websites and unreliable online information because important health decisions are frequently made based on online information. Healia and a few other search engines have been focusing on this issue for several years. As Erick alludes, the challenge is to find methods that are sensitive and specific, scalable, and minimize the gaming of search results.

  • Took a search of “Ludwig Wittgenstein”. The credible search result was a invalid page. Damn!

  • algore search engines are dead.

    LocationExpert.com

  • This feature is good if Hakia is able to implement it properly and effectively. The current beta version is offering credible sites only for Health and Environment topics.

  • Web 1.0 was about read-only and moving traditional applications to the Web. Web 2.0 is about two-way read-write and social media. And Web 3.0 is all about the semantic Web.

    Interesting thing is, enterprise search and other content management systems have taken quantum steps toward fulfilling the promise of knowledge management. With the rise of Social Knowledge Networks, both vetted information and social content can now come together to bring only the most relevant information to a collaborative community.

    Are Social Knowledge Networks the first true step toward Web 3.0 and the semantic Web? It might be a matter of “semantics” but one thing is clear … this is where the Web gets interesting…

  • I really like the Hakia’s tabs interface. It will be nice if they can also extend their index.
    The credible sites feature seem very important these days as I see many spamming web sites. Usually most of these sites don’t provide any valuable information, they are just optimized for search engines.

Leave Comment

Commenting Options

Enter your personal information to the left, or sign in with your Facebook account by clicking the button below.

Alternatively, you can create an avatar that will appear whenever you leave a comment on a Gravatar-enabled blog.

Trackback URL
bugbugbug
Techcrunch on Facebook