-Or

I developed this website mostly during my summer break. I do some improvements now and then when the idea is interesting enough. The project started from my research on social tagging a.k.a. folksonomy.I found out that the sum of tags gives very good description of the website. It even can distinguish fine differences between websites.For example, the website that has the most similar tags to http://www.cnn.com is not usatoday.com or even cbsnews.com. It is msnbnc.msn.com. I didn’t think it was an accident.This motivated me develop a website that can help people find similar websites. Although I am a decent programmer, I had to learn php, css, ajax and jquery during the development process. The part that gave me the biggest headache was the interface design. It is more difficult than it seems to be.

As you said, there is room for improvements for similar-site.com. That is exactly right. I have quite a few ideas that could improve the service significantly and even expand the scope of application. But it takes efforts and investment. So far I have spent less than $200 on domain names and hosing. I am looking for collaborators on this project. Hopefully, we can make this website a really useful one.

When I was reading Dan Brown’s new book, The Lost Symbol, last night, I was intrigued by the discussion of noetic science in Ch. 15. I thought it was a fiction, but the author mentioned a website: theintetionexperiment.com. I immediately opened my laptop and typed the URL into the search box of similar-site.com. To my surprise, it gave me a list of results with high confidence. One of the results is http://www.noetic.org. I did a similar site search to http://www.noetic.org (it has Dan Brown’s book in main page), it gave me a list of websites that included all of the research projects mentioned by Dan Brown, including IONS, PEAR, and theintentionexperiment and more not included in his novel. I was quite delighted by the results. Probably Dan Brown already did thorough research for the book, but I still hope he could check out my website :)

I tried these websites today on the other two sites that provide similar site search you mentioned in this article. They did not return IONS, PEAR and many other websites. Unlike keyword based search, similar site search really depends on what users want, which may not be exactly what the searching algorithm gives to users. For example, google.ru is a google’s russian search engine. Probably, most users are looking for other search engines in russian. However, google.com, google.ca,…… are also indeed similar to google.ru. I found out that the other two sites returned a few russian search engines. It is, however, possible to configure similar-site.com to exclude ‘google’tag from the search which will return even more russian search engines than the other two. Unfortunately, people rarely use this feature. I am writing too much for a comment. I will stop here :)

On the other hand similarweb needs a couple more functions to give me what Xmarks gives.

I think Xmarks wins for me hands down. I have tried all three and that is the one which makes most sense to me. I think what we need is uber-plugins that do more than just one thing. Xmarks is a good example of such a service.

The abilities SimilarWeb is giving me on finding similar sites, similar articles and tweets with the Xmarks bookmarking is making my life easier.