There is a new search engine in town called Yebol, a search engine that does manage to combine various search features in a way that sets it apart from its competitors.

Yebol has developed a technology that combines the advantages of a traditional Google like search engine (spider the web, and add an algorithm that determines what’s “best”) with a human touch (group high quality web sites into different categories that may be relevant to the query).

The clustering part does have a human component — i.e. there are flesh and blood human beings reading and categorizing web sites and adding them to the database — but most of it is based on computer based algorithms.

Yebol puts it this way: “We have developed a group of algorithms of association, clustering and categorization for automatically generating knowledge for search concepts, web sites, web pages and users.”

Other clustering search engines

Yebol is not the only search engine in town that provides clustering, i.e. that sorts search results into different categories depending on the alternative meanings of a search query.

Vivisimo’s meta search engine Clusty has been doing that for a long time. Microsoft’s Bing and Ask also give alternative search categories or queries in the left hand column.

A set of boxed search results and alternative categories

Yebol, however, takes this concept much further, in that the whole search engine result page is presents different types of clustering.

For some 10 million predefined queries Yebol will provide you with a search engine result page that gives you a set of boxes containing search results and categories sorted according to different criteria:

1. There is list of related topics in the left hand column (cf. Bing and Clusty).
2. There are direct links to the relevant topic pages on authority sites at the top of the center column. (Yahoo! does something similar on a much more limited scale.)
3. There is a list of subcategories in the right hand column, again with direct links to relevant authority sites.
   The right hand column also has a section with expanded searches, which simply rephrases your search to a more precise query. Again there are direct links to authority sites.

4. The right hand column also has a box containing relevant twits from twitter. It has a scroll bar to reveal even more entries.
5. There is a news section in the center column.
6. There is an old-fashioned list of search results at the bottom of the center column. This is the box you use when everything else fails, according to Yebol.
7. There are also two clearly labeled boxes with “Sponsored results”, i.e. text ads.

Combining algorithms with human input

For some searches there will be additional boxes. Searches for famous persons will, for instance, give you boxes containing images and video screenshots.

In addition to this Yebol will give you alternative search queries as you write your keywords into the search form (as do many other search engines, including Bing, Yahoo! and Google).

Finally, you may also refine your search by clicking on one of the links at the very top of the page: “News” for news search, “Images” for image results, “Video” for movies and the default alternative called Semantics (they should find a better word for that one…).

The only alternative we are truly missing is Blog search.

10 million predefined concepts

The 10 million predefined concepts (which are stored by Amazon.com by the way — Yebol is using their “computing in the cloud” service) will be able to lead you to some 1 billion web pages, Yebol says.

However, the technology is scalable. They claim that they should be able to build a knowledge base of 100 million concepts over 10 billion web pages in the next phase, after they go out of beta.

Yebol is good at what it does

Does it work?

Yes, it does. We tested a number of medical terms, and Yebol continued to provide us with relevant links to sites like Wikipedia, About.com, MedlinePlus, Web MD, Mayo Clinic and so on.

For an obscure search like renaissance scientist and magician Giordano Bruno we got fewer results. But still: The links to sites like Wikipedia and Renaissance Astrology (!) were relevant, as were the left hand list of characters from Religion and Science. In that column there were also links to Frances Yates, the number one scholar on Bruno — so we were happy.

When we did a search for the Norwegian Prime Minister Jens Stoltenberg we found ourselves outside the 10 million as there were no sub categories or expanded searches. That being said, Yebol did give us lists of Norwegian Labour Party politicians and government ministries of Norway. All relevant, as far as we see it.

And yes, we did a search for Barack Obama as well. He was definitely included in the core 10 million, and included direct links to photos, videos, news and — interestingly enough — John McCain.

We are actually very impressed with the quality of search results. The way Yebol organises the boxes is fine with us.

We are particularly pleased with Related topics and Expanded searches. The first one is great for doing in depth studies of a topic. The second helps you define clearly what you are looking for.

Is it too complex for the regular searcher?

The question we always ask when we look at a newcomer like this one, is whether the search engine has any chance of succeeding in the race against Google, Yahoo! and Bing.

It has been a long time, indeed, since any new search engine became part of the leading pack. Much hyped search engines like Ciul and Wikia Search have made little headway. Wikia Search is no more.

For a search engine to succeed it has to provide a search experience that truly helps searchers find what they are looking for, and in a way they find to be much better than than the one provided by their current favorite search engine (which is, in most cases, Google).

It has struck us that for people like ourselves (who know a bit about the various “Top Sites” that Yebol lists), Yebol will actually provide a better search result interface than Google. We know that the Mayo Clinic may provide more trustworthy medical information than the Wikipedia, and will click on that link for a search on “Asthma”.

We also appreciate the fact that Yebol helps us identify quality information sites that we did not know about.

But if you do not know much about sites like these, you will have to use Yebol for a while in order to learn how to identify the best and most relevant for a query. Yebol only gets you so far when it comes to narrowing down the number of trustworthy sites.

Searchers that are looking for a quick and dirty solution may not have the patience to do this. Moreover, many may be confused by all the boxes and the alternative options given on the search result page, mostly because they do not have the search competences needed to make sense of it all.

Google keeps it simple for a reason

The reason Google continues to stick to its “one list of search engine result model” is not that they are stupid. It is because their various user tests tells them that a majority of searchers like it this way. So, instead of adding a box with video results, they include selected videos in the regular listings.

Still, there is hope for Yebol. There are now so many Internet users out there, that getting even a sufficient number of internet and information savvy searchers on board may be enough for a sustainable business model.

Moreover, Hongfeng Yin, Ph.D., the founder of Yebol, worked for more than four years at Yahoo! Data Mining Research team, where he built the core behavioral targeting technologies, products and more than a thousand models. Maybe Yahoo!, or one of the other major players, will consider buying Yebol when they see that it is working.

Here’s how Yebol presents itself:

See also AltSearchEngines report on Yebol