Google and the future of search

I read an interesting article today about JC Penney gaming Google. How did Google find out? They had to be told by a journalist. So who looks worse, JC Penney or Google? In my opinion Google does. JC Penney did what many business do–they exploited a loophole. And what they did wasn’t illegal. They just pissed a company off–the wrong company mind you.

So, Google’s algorithm isn’t doing what it is supposed it. It got gamed and it constantly gets gamed. So it makes one wonder how far can a algorithm go before it stops being valuable because it is constantly outsmarted by humans. Humans who don’t even know the recipe.

Google is obviously great at what they do, but it can be better. Daily we search and find things that aren’t relevant and Google takes that into account as best they can–like a click through to those links and bounce rates. But is that enough? A click through is the equivalent of going through a turnstile–just because you go in doesn’t mean you enjoyed your visit. And just because you hit the back button back into Google doesn’t mean you didn’t like what you read. I know for instance when I’m searching I will right click a link and open multiple versions of the items that show up. I’ll then go to those pages to see what answers my question. Google has no real way of knowing which satisfied me the most because more often than not those are one click and I’m gone because I was either satisfied or not satisfied. Sure, if I continue to search for the topic that may signal that I’m less satisfied–but maybe I was partially satisfied.

So when does Google take into account real human interaction? They want us to apply interaction to photos on Picasa by tagging people. Why not allow for a system of ratings for search? Why not tag items in Chrome? Why not prompt users and ask if something they are reading in Chrome is relevant to the searches that are coming up for those pages–regardless of how you got there. And if something is spam then it shouldn’t even show up–but that still continues to happen.

Sure, it could be abused like all things, but if they are a Gmail user or have a Google account then that would limit such issues as well as tracking via a Chrome/computer combination. Machine based learning has been good for Google–but it still has a ways to go and maybe that last 80-100% cannot be overcome with any meaningful way and a huge cost, but wouldn’t it be great to know a million people liked a page based on your same search query?

Online learning–the good and the bad of free

It seems really obvious: You get a free education and that is great. Stanford offers some of their Computer Science courses free–I’ve worked through the first and discussed this on my blog. It has been fun for about two weeks, but now I’m realizing a few things:

1. As the class gets harder there is no professor to discuss issues.
2. I’m putting in a lot of work for no tangible payment. Sure, you learn something, like the programming language Java, but you can’t exactly do more than impress your friends with your new found talents. You don’t earn credits towards a degree or certification. You can’t put it on your resume. So its time spent without building toward anything.

So, is this just complaining about something free? Yes.

Which is why I decided to find an online Java course and get actual credit that could be used towards a master’s degree. I’ve looked at several universities–such as Illionis and Colorado State, two examples, that offer online graduate degrees in Computer Science. but I need the prerequisites first. And that is where my local community colleges came into play. Asnuntuck Community College offers Object Oriented Programming with Java online which is perfect for someone who has a job and can’t make it to campus. I’m taking the course now and couldn’t be more excited.

I get all the great features of Stanford but with credit hours–and the cost of a few hundred bucks. Not only that, the fact that I have grades and cash on the line I’m far more motivated to get the work done in an organized and timely manner.

CS106a: Lecture #3 and #4

Lecture three ends with the entire Karel book and Lecture four jumps right into Java programming with “The Art of Java”. I just got done reading Chapter 3 and will watch the video for that lecture tonight. I have to admit, its a big jump from Karel and the first two chapters to chapter three. I’m still working on the final two Karel homework problems. I will finish those in the morning and report back on how that goes.

Stanford CS106a Lecture #2

Lecture 2 was great. If you do all the readings (chapter 1-3 of Karel) you will be way ahead of the game.

So far I have attempted the first assignment and got Karel to actually work. This is the one issue with doing this for no credits and on your own–no feedback from a professor. You never know if you know what the heck you are doing.

Day One Lecture

Mehran Sahami gives introductory information and housekeeping, but he keeps it interesting regardless.

The reality is that you could skip the first lecture altogether if you are just taking the class like I am.

Day One: Stanford’s Introduction to computer science | Programming Methodology

Why Stanford over MIT? Stanford’s course seems to be the most accessible to anyone. In fact its a course that most students take at Stanford regardless of their major. MIT’s intro course jumps right into coding with Python and is probably a bit more advanced and not where most people would want to start.

Here is the course

How am I keeping myself organized? I’m creating Google Docs and folders where I’m keeping my notes and PDF uploads from the course–which includes the textbooks as PDFs when available. If anyone wants to see this I may make it public.

The next post will be post lecture one.

Learning Computer Science online free

Having worked online for thirteen or so years and having some development skills, it occurred to me that it has become increasingly important to get a hardcore computer science background. I have a business degree and a nearly completed masters in journalism however–the world is becoming a place of haves and have nots. Haves are those who can code. And the have nots are the rest of us.

So over the course of the next year or so I will embark on earning a degree in computer science for free–without the official degree. Even better these will be earned at world-renowned universities like MIT, Stanford and Harvard. The great thing about computer science is that if you can build amazing things the degree is really secondary. After all, neither Bill Gates or Mark Zuckerberg have computer science degrees.

This blog will continue with other observations, but the main goal will be to build an online degree in computer science for free and tell you how to do the same.

The resources:
MIT offers free online computer science courses.

Stanford offers their engineering everywhere with multiple computer science courses.

There is another great list with free courses from multiple universities, but they aren’t in order of how you should learn.

I’ll start this process on January 3 and spend at least one hour a day working through the coursework. Tomorrow the process begins with finding which course will be the best place to start. I will not work through the process like a normal university. Instead it will be an hour a day during the week with no mid-term vacation or a few hours a day only three days a week. I will do my best to describe most of the work as it comes along and the things that work or tripped me up and how I worked around the roadblocks.

Web 3.0: Google Chrome Web Store

The Google Chrome Web Store is pretty cool. But the most amazing thing is the Huffington Post and New York Times web site apps. It’s a great way to view a website. New York Times is more conducive to reading articles than HuffPo, but HuffPo has a cool interface and galleries. There is nothing cooler than seeing new technology take on old media like New York Times and see another site redeveloped in a new format. Once they start adding interaction and other features we might be looking at Web 3.0

50% off

I passed a store today where it has a banner hanging out front that said 50% off sale every day. Is it really a sale if it happens every day and everything is always 50% off?

Google Analytics training can help you to love numbers

Lots of people hate numbers, but they would love them if only it was obvious how much they could help. If you have a website and don’t have some analytics program on it then why do you have a website? If it only servers to send photos to your family or its a closed blog then it makes sense, however if you are trying to sell something, build an audience or do anything that involves getting someone to your site then you need to know that they came.

Google Analytics can be daunting. There is no doubt. Omniture can be even more so. But there are so many resources out there that it shouldn’t be something to fear forever. You can pay to get a hyper training seminars in three days by Google Certified and endorsed trainers or take the Google Analytics video classes. I prefer the book written by Google’s resident analytics guru a Avinash Kaushik. href=”http://www.amazon.com/gp/product/0470529393?ie=UTF8&tag=hiponline02&linkCode=as2&camp=1789&creative=390957&creativeASIN=0470529393″>Web Analytics 2.0: The Art of Online Accountability and Science of Customer Centricity

As good as Kaushhik’s book is you probably want a step by step so buy this book for a great primer: Advanced Web Metrics with Google Analytics, 2nd Edition.

Don’t fear Google Analytics. Create an account and get started. There are too many great resources, many free, to put it off.