Semantic Search Engine, Yebol Set For November Launch


Startup Yebol promises unlimited semantic search as Google and Bing alternative, but will its ‘human knowledge’ engine be a differentiator?

Another search startup plans to challenge Google by offering semantic search results, that have been disambiguated to be more relevant, for every term imaginable.

Launched as a beta in July, Yebol clusters and categorises search terms, websites, pages and contents, rather than offering the standard list of web search results that the likes of Google, Yahoo and Microsoft return in response to search queries.

Also, while existing semantic search engines such as Hakia and Kosmix rely on language patterns to return results, Yebol’s results are powered by “human knowledge,” providing what the company argues are more accurate and relevant results.

The company currently indexes 10 million semantic search terms, compared with Microsoft Bing’s six million semantic search terms.

When Yebol launches in towards the end of this year, it will provide semantic search for every conceivable search term, something no company has done before, Yebol spokesperson Grant Landis told eWEEK recently. Moreover, when someone enters a term Yebol has never seen before into the search engine, Yebol will index the term in real time and produce complete semantic results.

The Yebol home page, which, like most search engines, includes tabs for videos, news and images, creates a customised home page for anything the user searches for. For example, a search for “pizza” on Yebol returns related topics on the left, including different types of pizza.

The centre rail is very busy, packing in not only search results for pizza, but also news and links to pages on pizza from resources such as Wikipedia, StumbleUpon and MapQuest. These sources are in the Top Sites box at the top of the page, and it is clear that Yebol is positioning the box and its sources as a competitive differentiator. The right-hand rail includes search options, as well as a real-time Twitter section that surfaces tweets that include the word ‘pizza’. And search on “pizza, Fairfield, Ct.” returned local pizzerias, along with a Yahoo Map offering directions to five restaurants.

A search for “cancer” shows a similar setup for results, but with a key change: A Categories section appears atop the Expanded Searches and Twitter boxes in the right-hand rail, offering to take users to Websites with information on cancer causes, treatments, statistics and more.

Google, Yahoo, Microsoft’s Bing and semantic search providers offer some of what Yebol aims to offer in total. For example, Google boasts great scale, but it does not rely on a bulk of semantic search to achieve it, nor does it surface Twitter tweets. Bing indexes Twitter tweets, and, like Hakia and Kosmix, it provides semantic search.

But none of them currently cover as much ground on the web from a semantic standpoint as Yebol will be able to do at its launch, Landis claimed. The company actually has all of the terms indexed, but is refining the visual presentation of them, he said. The real-time capabilities ensure that Yebol’s index will grow at a prodigious rate.

When it comes to other search engines: “They’ve all got X number of search terms, usually one and two-word search terms that they have been able to create algorithms for that scour the internet for related topics and give you additional information for that search term outside of just the PageRank lists [from Google],” Landis said.

Yet Landis claimed that is no longer how users are searching. For example, before users became so reliant on search engines like Google, they conducted basic one or two-word searches such as “Tom Cruise”. Today, Landis argued, users search for more complex queries, such as “Tom Cruise Katie Holmes fight.”

But this all depends on context. While it is true that users are conducting more complex searches on Google, Bing and other engines, there are still plenty of basic keyword searches piping through search engines. Indeed, a user looking for biographical information on the actor may likely enter “Tom Cruise”.

Even so, Landis said Yebol will provide semantic search results for searches that span entire paragraphs in length. Yet even if Yebol pulls this off, it has no well-lighted path to success, as few search engine startups prosper in Google’s massive shadow.

For example, Cuil launched in 2008 but has hardly made a dent. Bing, thanks to Microsoft’s brand and $100 million marketing push, is gaining some market share although, at 8.9 percent, it is well behind Google’s 65 to 70 percent share of the market.