Well, there is nothing specific for India as far as Google News is concerned. But within English speaking countries, for example there are a few specifics. We take the same brews but emphasis the news for that country. Local news is a given importance, local newspapers would are also given importance. For instance, in sports, Indian newspapers would cover local interest better. Also, we include lots of international news which would have the country’s name, which also adds up to local interest. We often try to make the lead story for Google News form the local clusters. The reason for that is they’re able to make sense to the local audience; they provide the right introduction to the audience. Also, a locally located cluster males it faster for the people to access it.
Also, Google News is available in certain local languages, like Korean for instance. As yet there are no local languages in India on which you can read Google News.
Sunday, February 11, 2007
Q. Was there a problem with the newspapers and other news souces when Google News finally launched?
No, not really. When the alpha version was first launched, it was a very modest offering. We didn’t have too many newspapers sources. Besides, a news search was not entirely a new concept. Google had been linking to their articles previously and newspapers were aware there was a steady stream of traffic from Google search to their web sites. Besides, on Google News, there is just a title and a small write-up on the article.
Q. Can you tell us a little about the ranking system and how the news was ordered?
Well, there gad to be some rationale behind the ranking system. So we used a couple of parameters- how fresh is the news and how diverse is the news in question. If people are writing many original articles over a topic and all of them have recently written, obviously that topic has excited editors. So, that would definitely be news that we would rank high. Over time, for every story, I could measure how fresh it was, and how much momentum there was in the news, this formed the basis for the story rank. We don’t link to news that was older than three days on Google News, but if you run a search you can get news that’s up to a month old.
Many sites allow us to crawl through their content, and if a subscription is required, we flag it. The number of hits doesn’t determine how stories are ranked; they’re ranked purely according to editorial interest. So, within a particular cluster, the stories are cycled and every newspaper gets to the top.
Besides, we’re making all the news accessible in just one place, but with various point of views. For instance, we had articles about the Iraq war not just from the US, but also from Iraq, Europe, and the rest of Asia. This broadens people’s mind by bringing in many perspectives rather than seeing just one side of the story.
Many sites allow us to crawl through their content, and if a subscription is required, we flag it. The number of hits doesn’t determine how stories are ranked; they’re ranked purely according to editorial interest. So, within a particular cluster, the stories are cycled and every newspaper gets to the top.
Besides, we’re making all the news accessible in just one place, but with various point of views. For instance, we had articles about the Iraq war not just from the US, but also from Iraq, Europe, and the rest of Asia. This broadens people’s mind by bringing in many perspectives rather than seeing just one side of the story.
Q. We want to know about Google History to Beta launch in 2002.
Reply: Well, Google's charter is to make all the information in the world accessible. News is a very important part of information. The thing with news is that it's different with different newspapers and publishers. Nobody knows who has published what. You can get old content on the web, but news is fresh. News has its own challenges, quite different from web content. Newspapers and their readers don’t know all the news that is printed. There was an opportunity for both readers and journalists to understand what was published and what was missed. At that time the leading news sites completely missed the fact that there were multiple news avenues and choices.
Now after 9/11, many people read a whole lot of news. Most of it was online, since you just couldn’t wait for the next day’s newspapers. So, while people searched the Internet for news, they would often find the same article that they had read previously. It became a tedious process to find news that came in from different sources.
I realized that all of this could be automated. In fact that was a rather modest first ambition that I had- to automate the process of news. We wanted to build something that could access every possible newspaper online, rather than having to search every time. Then there was a generalist push- we decided to get all the topics all the time. This was by a process that we called “clustering.” We would try and get as much different news (through Google’s web crawlers) for a particular topic such as sports or politics that we called a cluster; I was able to get everything grouped by topic. Somebody pointed out that it looked like a newspaper, so we laid it out like one, adding more topics….
The demo of Google News was finally up by December 2001, three months after the project was conceptualized. In the beginning, we used news from 150 sources, and now we carry news from over 4,500 newspapers in the world. All newspapers have original content and we try to see that duplicated articles are avoided, using the story rank index. We don’t normally use wire service news, though newspapers do carry wire services reports at times. The first alpha was available on Google labs’ web suite from March 2002, and the current version was first up in September 2002.
Now after 9/11, many people read a whole lot of news. Most of it was online, since you just couldn’t wait for the next day’s newspapers. So, while people searched the Internet for news, they would often find the same article that they had read previously. It became a tedious process to find news that came in from different sources.
I realized that all of this could be automated. In fact that was a rather modest first ambition that I had- to automate the process of news. We wanted to build something that could access every possible newspaper online, rather than having to search every time. Then there was a generalist push- we decided to get all the topics all the time. This was by a process that we called “clustering.” We would try and get as much different news (through Google’s web crawlers) for a particular topic such as sports or politics that we called a cluster; I was able to get everything grouped by topic. Somebody pointed out that it looked like a newspaper, so we laid it out like one, adding more topics….
The demo of Google News was finally up by December 2001, three months after the project was conceptualized. In the beginning, we used news from 150 sources, and now we carry news from over 4,500 newspapers in the world. All newspapers have original content and we try to see that duplicated articles are avoided, using the story rank index. We don’t normally use wire service news, though newspapers do carry wire services reports at times. The first alpha was available on Google labs’ web suite from March 2002, and the current version was first up in September 2002.
Google Latest News.
An interview with Krishna Bharat, Principle Scientist Google
He is a quite busy man, but we set out to interview him on phone. As he was not in hurry so he talked with us for more than 45 minutes.
About Krishna Bharat:
He was born in Banglore, he graduated from IIT in 1191 from chennai with B.Tech in computers. After that he did PhD from Georgia Tech in 1996, then he joined Digital Equipment Corp research centre in Palo Alto. During his time there, he worked closely with Altavista on Search Engine Technology. After a few meetings with Larry Page and Sergey Brin, he finally joined Google in 1999. If you can don't know much about him then read more.
He is a quite busy man, but we set out to interview him on phone. As he was not in hurry so he talked with us for more than 45 minutes.
About Krishna Bharat:
He was born in Banglore, he graduated from IIT in 1191 from chennai with B.Tech in computers. After that he did PhD from Georgia Tech in 1996, then he joined Digital Equipment Corp research centre in Palo Alto. During his time there, he worked closely with Altavista on Search Engine Technology. After a few meetings with Larry Page and Sergey Brin, he finally joined Google in 1999. If you can don't know much about him then read more.
Subscribe to:
Posts (Atom)
