Hakia Review

Hakia: Relevant Semantic Search

As a technology writer for several of the world's most "in tune" blogs I have had the pleasure of reviewing many innovative websites and startup companies. One of my favorite innovations has been the hakia natural language search engine. Dr. Riza Berkan and an elite team of scientists and programmers have set out to create the world's finest and most relevant search platform. The task is a daunting one, but from interviews and research into the heart of this undertaking it is becoming obvious that hakia may well develop something a bit more than just Google with context.

Hakia is currently in beta testing but the site is available to the Web public with some of the explanation of their endeavors apparent in the hakia labs section. One of the most difficult aspects of new ttechnology is the "explanation" of the ttechnology. This is particularly true for hakia as the heart of the engine is comprised of advanced mathematics, linguistics, semantics, artificial intelligence and fuzzy logic. To the layman, even the most simplified explanations of these things proves difficult at best.

Semantic search is essentially breaking down queries into natural language the way yo and I have conversations. The degree of relevancy inherent in person to person communication far exceeds what Google or Yahoo engines can reflect. So, making a machine that responds to inquiry and returns "human-like" results is the task at hand.

Google by comparison uses algorithms to sort out word or term queries by using an index that is relatively planar or one two dimensional. From this index relevance is determined by a number of factors including page links, meta tags and keywords. As an example: a single 10 world sentence typed into the Google search bar could return millions of permutations based on an algorithmic model. Hakia, on the other hand breaks this same sentence down as a human would and rationalizes the results to reduce the permutations down to perhaps a hundred or more. As you can see, the semantic model is inherently geometrically more effective at returning relevant results.

The fly in the ointment is making an engine that can think the way humans do and then return the relevant results. Hakia is accomplishing this with language and mathematics in conjunction with what can be termed as AI. Hakia's indexing system is far more diverse and effective than Google's in that it responds more quickly and is not constrained to a linear index system. In effect, the Q-dex system emulates the cells of the human brain cells. Combined with advanced algorithms and fuzzy logic, this system will basically understand a quiery as a human does.

Finally, once the engine is online and available people will not only be able to do simple searches, but they will be able to effectively communicate with the "engine" in a much more intuitive and meaningful way. In theory hakia will be able to understand what a question is and be able to provide a set of "organized" and extremely relevant responses to that question. The implications are extreme and certainly that task is not in any way simple. Soon however a person will be able to search the Web without SEO tactics or spam reaching their desktop unless they desire it. Hakia has the best chance of meeting this task of the hundreds of web sites and dozens of search engines I have reviewed.