I remember reading somewhere online about a recent interview
(amongst many) with Ray Kurzweil – Director of Engineering at Google and the
world’s staunchest AI visionary – where he mentioned something about his plans in the line of…

to get the Google computers to understand natural language, not just do search and
answer questions
based on links and words, but actually understand the semantic content

I cannot remember where I saw the interview and will Google it…

…but I have a problem…

I know my keywords are fuzzy and I use Gatfol to show me some sample permutations on my input
…and realise that with 13 keywords and only 10 same-meaning single word replacements for each
keyword, as well as allowing for multiword to semantic equivalent multiword replacements, I am
looking at a minimum of 10 000 000 000 000 000 000 000 000 different ways to phrase my input
…all with the same basic meaning…

I also realise that I want a very specific search return that semantically equates to my input and
does not just hit and return pages corresponding to the trillions of different keyword combinations…

I quickly calculate that just by browsing my own thoughts and looking at my immediate
physical surroundings, I can theoretically come up with trillions upon trillions of meaningful
search input queries…each with trillions upon trillions of semantically equivalent phrasings…

The thought hits me that if Google wants to morph into Ray Kurzweil’s vision of a search
mind with which I can communicate and interrogate as if human – perhaps in five to ten years -
it will have to perform massively scalable multiword to multiword semantic equivalence transformations…

…Gatfol currently does this…

 

Can you see the future ?



Amit Singhal: Senior Vice President (Google Search)
…SXSW conference 2013, Austin Texas USA…

“having humanity’s knowledge on the Web is not enough, you have to understand it”

“search challenges: knowledge graph, speech recognition and natural language”

 ”…the perfect search engine should know
exactly what you need and give you what you want”

“our dream is to become the star trek computer”

“people have completely come to expect search engines to work and
the questions they have been asking have gotten harder and harder”

“…voice is clearly crucial for the future”

“…a star trek computer you could type or talk
and connect to would be crucial to (search) success”

 …all good and well, but why can we easily generate millions
of semantically equivalent search query replacements while
Google still cannot return semantically equivalent result groupings?

If I ask meaning wise the same thing – just with different
keywords – shouldn’t my return results be semantically consistent?

Why do we have to laboriously guess, replace and aggregate
synonyms as keyword alternatives to improve returned search results?

“….I would really like to find the web meeting discussion where the
participants discussed help and assistance to Indian farmers to help ease
yield decreases and with immediate concern funds and finance issues on the ground….”

AND

 “…I would actually like to get the web talk where the
participants discussed support to Indian farmers to improve yield
net gains and with the critical thought of money issues on the ground…”

 

 Not one of the top ten Google return
results for my first query is replicated in even
the top
hundred results of my semantically equivalent second search query…

 

We dream of talking to Star Trek computers but are still
stuck for optimisation in the “caveman-speak”
three word input search query universe…

 


Beam me up Gatfol…

 

 

I need relaxation and I know the right music will do it for me.

TuneIn.com offers 70 000 local and global radio stations and 2 million on-demand streams to choose from.

TuneIn also offers a search facility, which I use with the keywords “reduce stress”.

These are my first ten radio stream search returns :

But I also notice that if I use the single keyword “relaxation”, I get a totally different set of first ten radio stream returns :

 I start to wonder what different returns “reduce stress”, “anxiety relief”, stress management”, “healing music”, “calm mood music”, “deep relaxation”, “music for meditation” and “easy listening” might bring.

Wouldn’t it be nice if my single “reduce stress” input search could instantly also be translated to all the above keyword sets and all of them used to retrieve my search return?

 Gatfol technology massively enhances each and
every searchable online database with thousands to
millions of additional parallel search inputs in milliseconds…


Gatfol is search relief…

 

Igor Vladislav Stanislavovich Rozhdestvenskij

 1914 National Geographic:

…where the drosky drivers wore padded coats that wrapped snugly around their

already bloated bodies, because the more well-fed and fatter these coachmen were, the more
prosperous they appeared and the more likely they were able to fetch a higher price for greater profit…

Gatfol acts as search entry filter to transform long, complicated
searches instantly into relevant, compact, result-optimised queries.

Gatfol creates thousands of stealth search queries – “padded assets”
– equal in grammar structure and semantic meaning to the original query:

“…we really need to concentrate our architectural arrangements around a
stratified private social environment of street-level relaxation – in synergy with
modern culinary spaces and intellectual knowledge- that includes streetscape culture…”

…Gatfol translates to…

 “…we need to focus our design on combining coffee shops with book stores and art galleries…”

 Gatfol replaces current static keyword-based search methods
with next-generation semantic transformation technology

Gatfol makes business information-flow simple…

  …Gatfol padded stealth…your profit…

 


Six months ago Gatfol showed that if people have technological options, they would
like to enter input queries into search engines with an average length of around 20 words….

Real world queries…

-What is a spade flat head with a worm like body that has black stripes down center of body and moves slow about 2 inches long and 1 centermetre wide the body was all the same length? -Is there a movie or video that has footage of the deep dark corners of the sea that doesnt have a whole lot going on just some cool looking fish doing their thing? -Why does an organism that lives on land have more complex mechanisms to maintain a stable internal environment than organisms that live in water? -There is this roomer going around that says that 7 people have already been eaten at ratanga by the crocidiles at crocidile gorge is it true? -What are the three terms that are indicative of what kind of food source an animal from which it derives it’s primary of energy supply? -Which of the following are adaptations to remaining afloat and controlling bouyancy found in many pelagic organisms except? -What is the asexual reproduction method conducted by starfish in which new organisms can grow from its sections? -What is the maze of passageways that carries proteins and other materials from one part of the cell to another? -What are some of the conditions or requirements that distributed database would be better for an organisation? -Explain what Welty means by the habit of love and tell why this habit might be compared to a worn path? -Why humans have a higher density of receptors for touch in some areas of skin than in other areas? -How do put together your bio-ball aquaium filter system i bought used use everything in a a box? -How does the body’s arrangement of nerve receptors for touch make the Braille system possible? -Why do the survival of the whole organism depends on the functioning of its individual cells? - What is the relationship between the carbon in the ocean and the carbon in the soil? -If the dna sequence is tacgttaccgagctagact then what is the sequence of the messenger rna? -The effect on ancient cultures and traditions of people of cutting the old-growth forest? -Why there is more species diversity in southern Florida than there is in northern Alaska? 

…and that just five and six word length search queries made up a paltry 12% of all searches…

Latest statistics in June 2012 shows worsening of these already disquieting figures:

Use of  “caveman-speak” one word queries have risen 19%.
In each of the categories of five words or longer, usage dropped by at least 10%.

Gatfol easily manages search queries exceeding 50 words…

…Gatfol bridges the web to the world…

 

In the 1960’s Big Data was a child – compact and sprightly, in the 80’s it grew up to be slim and manageable. Today it is an unsightly massive volume – with the handling thereof – frankly grotesque.

Added bulk with the passing of time…

Roughly ten and a half gigs of data are being transferred each and every second…

Every day, we create 2.5 quintillion bytes of data…

Last year humans created 1.8 zettabytes of data…equivalent to one person taking 47 million years to watch 200 billion movies of at least 120 minutes long…

90% of all the world’s data has been created in the past two years…

The world’s technological capacity to store information has roughly doubled every 40 months since the 1980’s…

…and what will the massive data stream be like 50 years from now?

Big Data comes from everywhere: ubiquitous information-sensing mobile devices, digital images and videos, social network posts, wireless sensor networks, climate sensors, e-commerce transactions…

Today, the number of networked devices is equal to the global population. By 2015, the number of networked devices will be twice the global population.

Today there are currently about 400 million devices connected to the Internet, mostly phones and computers.

By 2020 some 50 billion devices, from cars to appliances, will be talking to one another.

Inevitably, Big Data is essential in helping to identify future business trends, eliminate fraud, filter harmful content, prevent life-threatening diseases and expose potential terrorist attacks…

We have to live with the big, but we can make it attractive…

 …effective search makes “big” beautiful again…


Let Gatfol get to the bottom of it and turn Aunt Helga into Megan Fox…

 

 

In Iron Man, Tony Stark (Downey Jr) tells a robotic assistant…
“you are of no benefit”
…and the robot moves out of the way…

We are actually light-years away from being able to get a machine to perform the action given this input.

None of the input phrase keywords – looked at individually – will effect the “stand down” response. Only when we parse our input into multiword groups – “you are” and “no benefit” – do we have a semantic language route to enable relevant robotic action.

We have half-a century of failed natural language AI repositories such as CYC, OpenCYC, Wordnet and MindPixel. Gatfol pursues a radical new approach strongly focusing on massive multiword-to-multiword input text conversion with semantic and grammatic integrity.

Gatfol develops the ubiquitous underlying technology supporting all future human-to-machine conversation, whether it be telling the airconditioner to switch to “low”, updating the GPS in the Ferrari to a new destination or even instructing our robotic arm to get the #$&%*? out of the way.

 

 

A housefly is the most accomplished aerodynamicist on earth.

It can make six turns a second, hover, hurtle and do somersaults with 90-degree reversals – all with a brain smaller than a sesame seed.

How ?

The secret is in the fly being able to process multiple inputs to simple output extremely quickly. Roughly two-thirds of a fly’s entire nervous system is devoted to processing large streams of incoming stimuli. If these multiple signals had to be routed to an approximate equal amount of output permutations, a fly would be bogged down with process management very quickly. Instead – the massive amounts of sensory data is boiled down to a few basic commands such as “left”, “right”, “up” etc.

In the same manner Gatfol processing is able to perform multiword semantic conversions to single words extremely quickly. But – in addition to the fly – it can do the reverse (single to multiple) equally fast. This enables Gatfol to pick up core concepts in large datasets and like a fly – produce focused output to optimise operational input.

Gatfol crystallizes simple concepts from vague, extremely complex or ambiguous data

Gatfol is applicable across all industries where search – or analysis is involved

 
 
How does Gatfol work algorithmically?

 

 Lets look at the following input phrase.

“…I am looking for a small coffee shop with red canopies in central
Copenhagen that serves many types of strawberry cheesecake…”


Gatfol builds a mathematical world image through web
spidering and “sees” keywords as negative semantic spaces


Gatfol combines negative spaces of groups of words i.e. “coffee shop”

Note how the small “coffee cup” negative space adds “coffee” to the full “coffee shop” world image

We do not have “coffee” as a concept, neither do we have “shop”
as a concept and NEITHER DO WE HAVE “coffee” and “shop” as a combinational concept

We have a unique negative space of  “coffeeshop” that’s neither individually
or in combination part of the input concepts

Gatfol can easily and fluidly expand the “negative spaces” to contain
almost any combination of concepts in an input query

Negative space templates are multi-dimensionally compared for the closest fit: giving us…

“….Maurice’s deli with the largest variety of strawberry confectionary,
in walking distance from Copernicus square…..”

Gatfol additionally provides an error protection layer:

Semantic fits are re-fed into the world image to ensure
grammar accuracy in the search phrase equivalent

 

 

It’s all very confusing!

 
How does Gatfol fit in with search engines like Google?…

Let’s take our favorite example as Google search input :

“….the small coffee shop with red canopies in Central Copenhagen
that serves five types of strawberry cheesecake….”

What Google Sees

What Google "Sees" (with quotation marks)


Gatfol “sees” much more…

The Gatfol and the Google query universes…

 

 Gatfol generates hundreds to thousands of stealth
replacement search queries to the original

 

Semantic Intelligence Filter Technology

 

 What is the underlying technology driving Gatfol’s semantic analysis engine?

Gatfol drops input search words through patented matrix “filters” each with a different “focus”.

Think of the landscape images below: a wide “focus” sees only a lake with mountains,
but a narrow focus additionally sees a country house in the landscape.


 

 


The future of search from the viewpoint of a large search engine like Google:

Google Search 1998

Google Search 2005

Google Search 2010

…Google core web search has not qualitatively advanced much over the years,
apart from substantial window dressing….


….the only quantum jump development in core search
HAS TO BE intelligent natural language interpretation…

 

 What Gatfol core technology will turn Google into….

 

Blog Home © 2014 Gatfol Home