For example, when you search digital cameras, it shows the images of 1,562 digital cameras in an semi-3D array, see image on the left. You can scroll up and down to pick the ones that look nice to you, or you can use the feature selector on the left to select number of pixels, zoom, price range, etc, and it will narrow down the image array to the ones matching the selected features. So it provides quick way to visually find a camera you like.

On Bing Visual Search, it also offers other pre-made visual searches, including dog breeds, handbags, movies, even politicians. The downside of this visual search feature is that each of these visual searches has to be manually built, very much like the old Yahoo directories, only now with visual navigation. So there is the scalability and timeliness issues. Visual search will work well on things where the looks matter, these actually include many things, such as handbags, shoes, sunglasses, furniture, home decor, fashion, travel, pets, etc. As long as a category is popular enough to warrant the man power to create a visual search for it, it can be done. Because it is pre-made, it will suffer in the real-time-ness of the data. New items may not be included days after launch or until it becomes popular.

Many players have tried different ways of doing visual search, and they all have failed to gain traction. The right areas for visual search should be things where looks matter, there is huge search volume, and it can be monetized.  Shopping and travel are two top candidates. Another area is dating. The first dating site that will bring the Bing Visual Search feature for browsing potential dates will be a hueg hit.

Bing has been providing good search results on par with Google. With the new visual search feature and other planned features to be relased around Christmas this year, we believe that Bing will start to gain market share. This is good thing because when there is only one dominant player in an area, the competition and hence the innovation goes down, and the prices go up.  However, people’s habits are hard to change. Most people’s search needs are met by Google, I doubt the visual search will be enough to convert many people. It may draw some curious drive-by traffic, but it may not be enough to draw people back when they need to do quick search.

Another example is the 8% discount you get on eBay buy-it-now when you use Bing to search for something and then follow the eBay ad to eBay. Many people first pick something they want to buy, just before they click the buy button, they go to Bing to search for a product, then click on the 8% off eBay ad to come back to eBay. Refreshing the screen of the item you want to buy, the 8% discount now shows up. The 8% discount draws people to Bing, but these same people may not use Bing to search for other things.

Surprisingly, one small feature in Bing that has drawn attention from school kids is the high definition images that Bing display on its home page, where when you float your cursors on different parts of the image, an interesting fact and a search link pops up.

What do you think of the search war between Google and Microsoft? What will make you switch to Bing?

Search engines, such as Google, which help us find useful information on the World Wide Web, are taken by granted by most of us. There are an astounding 150 million+ websites on the internet currently, according to Netcraft. So, how exactly does Google sift through this astounding number of websites, and find information relevant to your query?

The answer is that Google uses algorithms, which are complex mathematical instructions that tell computers how to carry out assigned tasks. When you type in a query, Google’s algorithm essentially searches out webpages that contain the keywords of your query. It then assigns a rank to each page, based on a number of factors ( e.g. how many times the keyword appears on the page).

The higher ranked pages are usually at the top of Google’s search engine results page (SERP). Therefore, you usually get the best links relating to your search query at the top of the list.

As Google is now not only the world’s most visited search engine site, but the most visited website itself (and also the worlds first $100 billion brand), being listed prominently on Google can result in a big boost in traffic for a website.

Google’s keyword search function is similar to other search engines- automated programs called spiders travel across the web, link to link, building up an index page, which contains certain keywords. This index is what is referred to, when a user enters a search query. Google lists the pages that contain the same keywords that are in your search terms.

A big part in how Google finds sites is played by Keyword placement. When Google looks for keywords in a webpage, any keyword in the webpages title or heading is given extra importance. Headings to come in ranges of different size, and the keywords in the larger headings are more valuable than if they were in smaller headings.

Google’s patented PageRank system

Arguably, the most important feature of the Google algorithm is the PageRank system. It is an automated process that determines where each search result appears on Google’s search engine results page. As most people only look at the first few search results, having a webpage at the top will be likely to generate a lot of traffic for a website.

And how are these results standings determined? Well, we don’t exactly know- the official algorithm is kept secret by Google. What we know do know is:

• Every search result is assigned a score by PageRank. The higher the score, the further up the search results will appear.
• Because pages with good quality content usually have lots of trackbacks (i.e. other webpages link to the target page), the scores are partially determined by the number of trackbacks a target page has.
• Trackbacks from higher webpages count more than the lower ranking sites.
• Other factors which affect the score are the longevity of the site, the strength of the domain name, the age of link going to and from the site. Sites that have been around for a while usually have more value placed onto them by Google.

This sort of strategy works quite well as, by focusing on links going to and from a webpage, the search engine can organise results in a very useful way. Arguably, the best way for your website to get a top spot, is to consistently churn out high quality content, as this will result in more webpages linking to you.

The Wolfram Alpha search engine will be launched this month, and it could possibly revolutionise the web. It is not the regular run-of-the-mill search engine, it is a new system of software which will be able to understand questions and then provide tailored answers. This will be the first search engine of its kind on the internet.

Don’t mistake Wolfram Alpha for Google. It is completely different. It is a store of information which understands and responds to language like any other person. It is what people (especially students!) have fantasised about.There was a sneak peek of the Wolfram Alpha, last week at Harvard University (Videos below)

Although the system is still new, experts believe that this is a massive leap in the development of the internet- that this is part of the natural evolution of the internet. Computer and Internet expert, Nova Spivack, suggested that Wolfram Alpha could be as important as Google:

“It is really impressive and significant…In fact it may be as important for the web (and the world) as Google, but for a different purpose.”

Wolfram Alpha will not only give a straight answer to almost any random question such as ‘How long is the Amazon river?’, but it would also neatly produce relatedand sourced information. Therefore in this case, it might also provide a list of towns nearby the Amazon river or tributaries running off the river.

However, the most impressive feature of the Wolfram Alpha is its ability, as its inventor Dr. Wolfram put it, ‘to work things out on the fly’. You are not limited to asking a simple, straightforward question. Ask it to compare the heights of the Empire State Building and the Petronas Towers, and it will do it. Ask it about Chile’s interest rates the day the Berlin Wall officially came down, it will cross-check and find the answer. If you imply a question within a statement, e.g ’20 dice rolls for five 6′s’, it will guess you want to know the probabilities of dice rolling and will figure it out.

One burning question might be- ‘Is this like Wikipedia? Is the information trustworthy?’ The answer is that it is not like Wikipedia in that sense, because the results that you receive are assessed by experts and are made available only after a thorough verification. So there is no need to doubt any information in the Wolfram Alpha.

Dr. Wolfram, 49, who got his pHD in particle physics when he was 20, adds that the launch of the ‘computational knowledge engine’ later this month is only the start of the project. However, it is already touted as the ‘Google Killer’, and has raked up plenty of hype before its proper launch. But with Google being one of the world’s top brands, worth $100 billion, maybe Wolfram Alpha has the potential to be one of the biggest names in the planet.

My view is that Wolfram Alpha can’t be a Google killer because frankly, Google doesn’t have anything like this (apart from maybe their new public data search). But in any case Google will lead the market for normal search engines. Wolfram Alpha is something different, I would see it more as Wikipedia’s competitor, than Google’s.

What are your views on the Wolfram Alpha? It is overhyped or is it proof of the ‘evolution of the internet’? And is it going to be a ‘Google Killer’?

Wolfram Alpha

A Sneak Preview of Wolfram Alpha (Excerpt):

Dr. Stephen Wolfram’s Talk in Harvard University (Full):