Personalised Content.. The next technology Milestone?

With the deluge of information available on the web and the little time that the peole have to wade through the corpus of information, personalised content will surely be the next breakthrough that the internet giants and the startups would be looking at..

Personalised news[ www.findory.com] , personalised search [www.google.com], personalised recommendations [www.netflix.com] and a lot more fields.

Greg Linden (the founder of findory) has written about it here
http://glinden.blogspot.com/2006/10/reddit-digg-and-personalized-news.html

The following article gives a market perspective of personalised news
http://www.readwriteweb.com/archives/personalized_news_market_overview.php

I was not aware that Yahoo had acquired Search Fox the pioneer in this field
http://www.techcrunch.com/2006/01/16/yahoo-acquires-searchfox-assets/

Yahoo has a huge potential for personalisation and behavioural targetting given their huge user base. With information about each user, they can easily profile him and give the user what he wants.

But the main hurdle in this field is generating correct recommendations inspite of sparse information about the user. The recommenation algorithms also needs to be scalable to churn out recommendations in real time.

I am aslo going through quite a few recommendation engine techniques, exploring some open source..
The ones that I found were interesting are

RACOFI
http://taste.sourceforge.net/
http://www.cs.waikato.ac.nz/ml/weka/

Right now I am on a vacation .. blogging from an internet cafe..

I will write about these tools [once I try them on the netflix data set] in the near future..

So get your thinking caps on.. and explore the ways to exploit this new technology and M A K E M O N E Y

Recommender Systems - Steering Wheels!

Recommendation systems are going to be the future of search.

Google with it's personalised search, www.findory.com, Amazon's Collaborative Filtering, Netflix Movie Recommendation Engine, http://racofi.elg.ca/ .. and many more


The fact that Google gives you a million results for your search is not a big thing...

[water water everywhere.. not a drop to drink!]

What matters is

• How relevant were those results
• Were the results the ones that you were looking for?

Recommendation engines will surely help us wade through the sea of information..


I am currently going through some of them..

http://www.nongnu.org/cofi/

http://glinden.blogspot.com

www.cs.waikato.ac.nz/ml/weka/

http://battellemedia.com/archives/000913.php

Personalised product recommendation in E-commerce by weng and Liu

Recommender Systems - Steering Wheels!

Recommendation systems are going to be the future of search.

Google with it's personalised search, www.findory.com, Amazon's Collaborative Filtering, Netflix Movie Recommendation Engine, http://racofi.elg.ca/ .. and many more


The fact that Google gives you a million results for your search is not a big thing...

[water water everywhere.. not a drop to drink!]

What matters is

] How relevant were those results

] Were the results the ones that you were looking for?


Recommendation engines will surely help us wade through the sea of information..


I am currently going through some of them..

http://www.nongnu.org/cofi/

http://glinden.blogspot.com

www.cs.waikato.ac.nz/ml/weka/

http://battellemedia.com/archives/000913.php

personalised product recommendation" in E-commerce by weng and Liu

Recommendation engine for Shops !

I am just back from buying some Sarees for my mom and some shirts for my dad..
Its diwali..

I had quite a trouble finding what my mom would actually like..
Firstly.. I dint have any domain knowledge [in sarees :)]
I dint know what type of saree my mom would like [Kanjeevaram,Bangalore Silk, Dharmavaram Silk, Mysore Printed, Crepe silk and some other ones]..
Wouldnt it be a good idea for the shops to record which type of saree is bought by what type of person?
Then we could mine the data to form associations [age, height, complexion, income group, type of saree, colour of the saree, the design on the saree, the time of the year the saree was bought...]
Then you just give the shopkeeper the profile and bingo.. You get the list of sarees that your mom [or girlfriend for that matter] would like..
This would be particularly useful when the subject is not present at the time of buying the saree...
Amazon does similar stuff.. but the recommendation it does it is for different objects..
Netflix does something similar to this..

May be we can have an online mart for sarees..
Where real models display the sarees that you would like [pre-recorded]..
We will have to do an analysis of the saree market in India..

Population on 1st March 2001 was 1027015247
Assuming that one third of them are women who would buy 1 saree each we would have 343 million sarees bought per year..
Even if 1 percent of these women are internet users [our target audience], we have a potential of getting
3.4 lakh customers.. Even if we make a profit of Rs.10 per saree we can earn 3.4 million Rs per annum

Okay Okay.. I know that this is not how it works.. I need to have a bplan, the data for recommendations and improve the above analysis[?] which is to rudimentary.. Maybe this is how I would explain the concept to my kid...

Gotta prepare for CAT.. got to give the Mock tomorrow :)

This summary is not available. Please click here to view the post.

My question on http://in.answers.yahoo.com on the first page of GOOGLE Search!!

When the Yahoo India Answers was launched I thought.. Just another Forum..
When I started using it I realised that it was something more..
Unlike forums which are specific to a certain subject (say CAT, GMAT, TOEFL, GRE, etc).. Answers was more like.. umm.. different.. It gave a feeling similar to asking your friend/neighbour/aunt for some help..
Yahoo is mainly thriving on it's huge community base..

Lets get to the point.. I am planning to buy an electric guitar. So I was seraching the net for some sources..
I was surprised to see my question on the first page of google search..
Is there a way to monetise this advantage.. My question.. on Yahoo.. top 10! thats something every SEO guy would look at!!

I tried pasting my Adsense snippet on the Answers website.. [Though I knew that it would not be possible :)]... It dint work..

..May be I can monetise this phenomenon by placing the Yahoo Answers Badge on My website..

I even took a screenshot of the page [couldnt upload it here though].. the query was
"electric guitar"+bangalore+buy

Content for a new Money Making website

I did a preliminary research on what a website content should be in order to get more traffic.
Most of the traffic that a site gets is through search engines GOOGLE, YAHOO. So if your site figures out in the top 10 (at least) results for the major queries, then your website would be a HIT. And you can have Ads on your website. Based upon this I came up with the following content

WHAT [SHOULD THE CONTENT BE]
-Most of the online users are from the US
Analyse using google trends

1. Celebrities
-----------------------------
http://buzz.yahoo.com/actors/
-----------------------------
Jessica Simpson
Beyonce Knowles
Paris Hilton
Scarlett Johansson
Britney Spears
Pamela Anderson
Shakira
Angelina Jolie
Carmen Electra
Hilary Duff
Christina Aguilera
Janet Jackson
Fergie
-Upcoming
2. Sports -
-----------------------------
http://buzz.yahoo.com/sports/
-----------------------------
NFL
WWE
Soccer
NBA BasketBall

3. Jobs
4. Comparitive Product Analysis/Reviews, cars, bikes
5. Tech
Linux
Windows
Wikipedia
6. Movies, Youtube also incressing
----------------------------
http://buzz.yahoo.com/movies/
-----------------------------
IMDB charts
-----------------------------
http://www.imdb.com/chart/
-----------------------------
Disney Channel
Cartoon network
Naruto
7. Recipe
8. News
9. Music
-----------------------------
http://buzz.yahoo.com/music/
-----------------------------
mp3

Now the Question is: Where do you get content for all this?
One way is [The difficult one] Go through various websites, read the relevant information and spit it out on your site.
Other way is (This is refered to as web scrapping) just pick the content from the top content providers.
One such good resource is WEB HARVEST (You need to know some XML and Java though). You can just tweak in the configuration file and scrap what all you want.

Finally this boils down to building a directory similar to YAHOO. The only difference is that you are arranging for your own content (Yahoo pays the content providers for this content!)

Make Money. START HERE
I have been a fan of Amit Verma's blog . http://www.indiauncut.blogspot.com/ for quite some time.
I had been reading it now and then for entertainment when I stumbled across a post which gave a list of top adsense earners. The post "From Google Adsense to Bling" which pointed to http://www.desipundit.com/2006/09/28/google-adsense-highest-earners/ was quite exciting.

From there on I started to explore the web for sources of making money.
Why do a job when you can earn money at home with minimal effort!

I then started researching to get some good idea which will fetch me good bucks.

In this blog I will document my journey towards online money making :)