The top 10 alternatives to Google search

Kosmix

Rather than returning a page full of the "infamous 10 blue links", Kosmix presents a web page that mimics that layout of a magazine article, offering facts, pictures, videos and comments on your chosen subject.

Kosmix taps into a raft of web 2.0 content, sucking in feeds from Wikipedia, Flickr and YouTube as well as blog and forum postings. It's definitely an interesting choice if searching for a topic rather than a particular website.

Image credit: Nick Heath/silicon.com

Cooliris

The prettiest of the bunch is Cooliris, a search engine designed for finding images and video.

Cooliris delivers results from popular photosharing sites such as Flickr and Picasa, video portals such as YouTube and Blinkx and online TV sites such as Hulu. It can even search for products on sites such as Amazon.

Scrolling through results is simply a case of flicking the 3D photo wall with the mouse, in a similar manner to the Cover Flow system used by Apple's iPhone.

Image credit: Nick Heath/silicon.com

Bing

When it comes to searching for sites, Bing offers a list of URLs that looks pretty similar to Google, as well as image, video, shopping and map search options.

However, Microsoft's latest Google killer also promises several features that Redmond would like you to believe will trump the search incumbent.

For certain types of searches, such as for products or hotels, it will display additional information in the results page, including user reviews and pricing.

Users can also refine results into categories and view video thumbnail previews.

Early signs are promising, with Bing pulling in 16 per cent of the global search market, putting it ahead of Yahoo!'s 10 per cent, according to internet monitoring company StatCounter.

Image credit: Nick Heath/silicon.com

Wolfram Alpha

Billed as a "computational knowledge engine" Wolfram Alpha is no simple site searcher.

Instead, it's designed to provide facts and figures, on anything from a country's gross domestic product to the frequencies of Christian names throughout history.

It processes more than 10 trillion pieces of data, producing results in the forms of data tables, definitions and graphs of comparative figures.

In its current form it is strong on answering scientific queries, providing numerical data and solving mathematical problems. It falls down however when asked for simpler everyday queries such as 'where's my nearest Chinese restaurant' or 'what's the best fertiliser for roses'.

Future plans for the engine include expanding its knowledge base to a wider range of everyday popular and cultural knowledge.

Image credit: Nick Heath/silicon.com

Hyperwords

Not a search engine per se but a plug-in for the Firefox and Flock browsers that effectively turns any word on a web page into a ready-to-use search term.

Simply highlight and right-click on a word to search a range of online stores, check the meaning of the word, find its location on a map, translate it into tens of different languages or even immediately post the highlighted text to your Twitter account or blog.

The embedded search engine runs queries against a range of popular sites such as Wikipedia, Amazon, eBay, Facebook and of course Google.

Image credit: Nick Heath/silicon.com

Mahalo

The self styled "human-powered" search relies on individuals using their grey matter to organise results and information, rather than the mathematical algorithms most search engines use.

The brainchild of self-styled web entrepreneur Jason Calacanis, Mahalo uses a team of editors to build pages of information around popular search terms such as the BBC.

Such pages offer a slew of well-organised and interesting content including encyclopaedia entries, videos, news stories, photos and facts.

For less mainstream content, no dedicated pages exist and instead Mahalo pulls in a mixture of URLs, pictures, comment entries and video, from a range of different sites including Flickr, YouTube and Twitter.

Image credit: Nick Heath/silicon.com

Cuil

Built by ex-Google engineers, Cuil says it not only searches more web pages than other engines but also returns more relevant results. It claims to do this by ranking search results based on a site's content and relevance to the search term, rather than just its popularity.

Results are broken down into categories at the side of the page and also feature short summaries and pictures.

Cuil isn't likely to have Google worried any time soon though: a search for silicon.com produces our site as the second result on the page, amid a lot of sites on the microprocessor ingredient and Silicon Valley.

Image credit: Nick Heath/silicon.com

Hunch

When you can't make up your mind about whether to ask for a raise or to buy your wife flowers then why not let Hunch do it for you?

Basically a tailored form of crowd sourcing, Hunch trains its search algorithms to answer future queries using users' responses to a series of multiple choice questions - helping it learn that people who are vegetarian will never choose a steak restaurant when dining out, for instance.

Unfortunately more training seems to be in order, as its recommendation for the best bunch of flowers for a loved one was a bouquet of scarlet pimpernels, commonly regarded as a weed.

But it's early days - the site was launched yesterday - and the machine learning tech should ensure it improves given more time and more users' answers.

Image credit: Nick Heath/silicon.com

Omgili

Omgili is the engine of choice for vanity searching or mining web communities for information: Omgili scours millions of debates on forums and discussion boards for mentions of a search term, be it a topic or person.

The engine even allows users to stream mentions of a search term as they happen in a real-time window, similar to updates on Twitter.

Omgili recommends using the site to find consumer opinions or to find answers to technical problems.

Image credit: Nick Heath/silicon.com