When
the average internet surfer can't find the news he wants, he simply
gives up. But when Krishna Bharat
faced that problem after New York's most famous twin towers
were reduced to a rubble by terrorists on September 11 in 2001,
he didn't click his browser shut in frustration. Instead, the computer
engineer from IIT Madras and Georgia Tech and Principal Scientist
at Google, set about creating a search engine that would not just
sort and group news stories, but automatically update them every
15 minutes. With that was born the now popular Google News, which
won the Webby Awards (internet's equivalent of the Academy Awards)
in the news category for 2003. The 34-year-old Bharat, son of an
IAS officer, also got Google its first patent, granted for a Hilltop
Algorithm he developed. Recently in Bangalore to set up Google's
India r&d centre, the soon-to-be-multi-millionaire (courtesy
Google's IPO) Bharat spoke to BT's Venkatesha
Babu. Excerpts from an exclusive interview:
The tale of how Google News was created
is now a part of the web folklore. Can we hear it one more time,
but from the horse's mouth?
Immediately after 9-11, I realised that it would
be useful to have news reports from multiple sources on any single
given topic assembled in one place. With thousands of competing
news sites on the internet, how can somebody get the latest news
about a subject in one single place? Nobody has the time to trawl
through the labyrinths of the internet to get informed coverage
from a wide range of perspectives. How do you manage a flood of
information to pick and choose what is relevant to you? It would
be time consuming to flip through one web page after another, scan
and then click on umpteen number of hyperlinks, and battle with
multiplying ads just to read a particular story.
Assembling all these under one single site
seemed sensible. But no single human being or even a batch of them
can do this job. The process of selecting stories and bunching them
together from credible sources could thus be best accomplished through
automation. So I wrote this algorithm that trawls the news websites
and clusters related stories. I felt this could be best done by
mathematics. The algorithm functions like a librarian or clipping
service, by searching out, matcing, and collecting articles based
on one's reading interest.
|
"Two things are key in Google News: Diversity
of opinion and impartiality, as it is completely automated'' |
That is how Google News came into being. Since
it is an automatic process, it updates every 15 minutes and does
this 24x7. When we go home, it does not mean Google News stops updating.
When we initially tested this in the immediate aftermath of 9-11,
the internal feedback I received was tremendous. Today, the site
is a news portal organised more like a newspaper updated continuously
as news websites post new stories or update them. We started off
with 20 sites. Today, Google News covers 4,500 credible news sources.
Google News is a force for democracy because of the diversity of
opinion it offers (on the same subject).
This was not your first attempt at organising
news, right?
Yes. Earlier even as a graduate student I was
interested in selection of news online and personalising it. At
the Krakatoa Chronicle, I had tried it. So, probably I was always
interested.
Even earlier, there were news search engines
like that of Alta Vista and...
None as comprehensive and none certainly the
way Google News is, nor (any) covering as many sources as we do.
The way you pitch it, it sounds like Google
News is the future of journalism.
(Laughs) It's presumptuous to say Google News
will be the future of journalism itself. It will be part of the
future of journalism.
News untouched by human hands. Should online
editors worry about their jobs?
That is not true. We are just a hub for news.
We are a content aggregator. Our technology trawls through the web
to find the most appropriate content and display it according to
relevance to the users. It was not intended to say that editors
don't have a role. We just assemble news from multiple sources.
News written by thousands of journalists like you. No single editor
or a bunch of editors could have done what our technology does.
You and other journalists are valued for the opinion, the perspective
that you bring to an issue. That does not change. Automation does
not mean that human editors become redundant; they just play a different
role. In fact, the search avoids bias by selecting all articles
that are determined relevant by the mathematical grouping, and not
just what a human editor would believe in.
|
"Larry and Sergey are aware of India's potential.
We will be doing R&D work in India'' |
If half of the web search traffic goes through
Google or your news site has a couple of million unique visitors,
you are playing the role of a gatekeeper in segregating news sources.
You have power. What is news to you and how do you decide who should
be in or out?
Google News is objectively created, not opinionated.
It reflects the view of many editors who create the content on the
sites Google trawls. Two things are key in Google News: Diversity
of opinion and impartiality, as it is completely automated. Relevance
is determined by information retrieval techniques that look at the
distribution of words in the article and surrounding pages on the
web.
Two years after it was launched for public
use (in September 2002), it still has a beta tag hanging on to it.
Why?
We at Google like to continuously tweak things
till we have perfected it. This is a work in progress. We continuously
seek feedback and based on it, improve things to make it more relevant
to its users.
Since you are a content aggregator, are
there legal issues if the site were to monetarily benefit from somebody's
else content? Has any news site objected?
Till now, we have not had this issue. Actually,
we might be driving traffic to sites, which a browser did not even
know existed. In fact, they might be benefiting because of Google
News (smiles). In one or two cases, when some of them have approached
us, we explained to them how we go about the entire process, and
they have gone back satisfied. If somebody wants us to remove them
from our list, we will consider it.
Any plans to monetise Google News or include
paid content like, say, a press release on a company?
I am only interested in technology. Once that
is done, it is not my concern anymore. There are other people within
Google who will be examining how Google might benefit out of the
work we do. However, at this point it is not something I lose sleep
over.
In several cases like the Wall Street Journal
or even Business Today, the website can be accessed either by a
subscriber or by a paying surfer. What happens when paid content
becomes the norm?
I don't think this will happen. The bulk of
news would still be available for free. But even if Business Today's
own site, for example, has a premium paid content section, as long
as they allow us to see and access it, they can then clearly charge
the end user whatever they deem fit and if he is ready to pay for
it. It is a win-win for everybody.
Tell us about your India plans.
Apart from our Japan, Zurich, and our Mountain
View centres, we wanted to set up an R&D centre in India. Larry
(Page) and Sergey (Brin) are aware of India's potential. A number
of people among Google's 1,900-odd employees are of Indian origin,
including me. We will be doing fundamental research and development
work in India. This decision to set up an R&D centre in India
was taken sometime in the latter half of last year. We were here
for a reconnaissance trip in December and after visiting various
cities and hearing pitches from state governments, we zeroed in
on Bangalore because of the talent available here. We have taken
a 15,000-sq ft facility in the city from which we will be operating.
The R&D centre will be on par with any
other in the world. Unlike some other companies, we will not use
Indian operations for low-end work like testing, quality assurance
or low-end programming. Our aim is that the next Google or Google
News must come out of India. The bar we have set for ourselves here
is very high. Some of us have been sent here to set up the India
development centre. Anybody recruited here if he or she does not
feel like working in India, can move to any of our centres and vice
versa. We are not looking at recruiting in hundreds. Currently,
we are just around a team of 20 or so, and we are looking for people
who are at the very cutting edge. We want creative thinkers, problem
solvers.
I believe Google plans to open another centre
in Hyderabad...
Yes, yes, we will be having a customer support
centre more like a BPO-type operation based out of Hyderabad. But
that is of a different type altogether. In Bangalore, we will be
doing basic research, not customer support. Unfortunately, we have
realised that there is a lot of hype surrounding Bangalore. While
there is a lot of talent available, we want the very best in our
category. We want the Indian R&D centre to be on par with the
best of the rest across the world.
|