For the readers that followed closely on TagEdge, you all know that I often profile on Israeli startups, and used to think that Israeli startups as not merely the creative technology firms, but also firms that will change the world eventually. A new social search engine called Delver is the one that I like indeed, my first impression was…eh, which was a too cold social search engine that not to sign-up and try.

Delver is currently in Alpha phase. If it is barely a search engine, or meta search, that’s not much I can talk about. However, it applied a new technology called social search which offer a way for users to bookmark, or Web, people, media search based on their social graph. If you don’t have any profiles in any social network, you won’t notice this wonderful feature. But if you register and link all your profiles, range from MySpace, LinkedIn, Flickr, to Facebook, you’ll realize the powerful search engine rendered by Delver.

From the profile linking, you can add your own search buddies, that enables you to give priority to results created or referenced by these people. Furthermore, you also can add social circles, which offer a way connect you with people that have something in common with you. Delver will take into consideration on who your friends are and it will return your search queries based the contents that your friends made on Delver as well as what bookmarks they’ve done in the past on Delver. Put simply, Delver can deliver a totally different search results based on your trusted friends you followed on Delver and indexed what they’ve done and their ultimate goal is to recommend your friends’ search results to you completely.

I’d like to share what I thought Delver is good in the practice that it allowed one to easily bookmark certain results with a single click namely “keep it,” the search result you like will instantly save to the right-hand sidebar for your future references. Also, I want to note here after played around it for a while, I knew that its ranking methodology is based on your own social connections, your friends’ quality contents will always be ranked higher than the other search results that based on the keywords you input into the Delver search bar.

In overall, I still feel Delver is very cool, but I’ll give it a second look, when it ready for the beta launch in the near future.

Yoozila is a new search engine that built on Google code. As stated on their “About” page, the presentation of the users’ search results in a beautiful Web 2.0 manner is their founding appeal. Since the foundation of their Web design is Web 2.0, you won’t surprise that the search result page is in 3-column, but I’m not sure when and how 3-column Web design has became a symbol of a Web 2.0 design and I also do not want to further argue about it.

Yoozila’s home page continue to follow the simplicity of user interface as preserved by Google. On the landing page of Yoozila, you’ll find the Web search, as well as search image, video, news, blog, and etc., seemed like it want to positioning itself as Google No. 2, but focused on Web 2.0 presentation manner. Most recently, it also updated a feature called word suggestion, which you can easily find a similar feature in almost all the notable search engines.

Using Yoozila for a while reminds me a lot of another search engine called Exalead. Both presentation of the search result page almost the same, I mean “almost.” Thumbnails on the left, search results stand on the middle, and right-hand side is the search filters. Differences are Exalead would run a sponsored advertisement on the top, and lack of the parental control which can be turned on or off by Yoozila users. Finally, both search engines also allowed users to choose their preferred viewing options (a small view icon on the top), that are to show image and title, or show title and text, or simply show it all.

Media Wombat is a Flash search engine. It crawled and indexed all the Flash and multimedia content on the Web into its database and you can find the Flash content you want by placing the keyword into its search bar. Although the Flash loading is known to be slowed, as we all know, however Flash’s potential in the multimedia space is still tremendous. There are many ActionScript developers who delivered well in their projects to their clients. Also, I believed the commitment of Adobe in this space, with their accomplished development work called Flash Lite focused on the use in mobile.

On Media Wombat, you can search the Flash content you want, with the filtering system of each result provided by Media Wombat. You can go for a Flash preview of a particular result, images or embedded on the Flash, as well as rate the Flash content or go directly to the result page through the link available. For the Flash or Flex developers, they can see the ActionScript on a particular Flash result simply by clicking the “View ActionScript” button.

Media Wombat is a project developed by Steve Webb and Troy Yohn.

Searchme is a visual search engine and currently in the public beta phase. Searchme reminds me a lot of another visual search engine called oSkope, however these two are totally different in terms of user experience. I like Searchme very much. It allowed me to search the stuff that I want and save it on my desktop, within a finger click, i.e. drag and drop the search results (in the format of page by page) to the stack I created. As told by Searchme, all the pages saved in the stacks are stored in your desktop computer, and if you deleted your Flash Player cache, the pages you saved will be gone.

There are other features in Searchme, you can delete the stack you created, or add a public URL to it. Don’t be afraid the pages you saved will be exposed to the public, good pages should be shared to others as this is an act of demonstrate trust and cultivate the knowledge management culture.

To know more, I embedded two videos of Searchme, watch and enjoy.

1) Introduction to Searchme

2) How To Create a Stack

Summize is a fastest Twitter search engine that I ever know. I find it especially useful when I want to track the buzz of a certain event or topic among all the Twitter users in nearly “real time.” After I keyed into some keywords into the search box for what I am looking for, Summize searched all the real-time conversations on Twitter and came back with the results, it showed “iphone” still one of the top topics among all the Twitter users, and I even see some tweets was indexed less than a minute.

Summize is a great service that compliment the Twitter Public Timeline, by providing a control mechanism on what you really want to read on Twitter. Will it be better to search Summize for the Twitter’s conversations when Twitter is down intermittently? Probably it is. By the way, if you like Summize, I think you’ll also like their soon-to-be launched Realtime Sentiment, Summize’s another project that currently still in labs.

Twingly is a new Swedish-based blog search engine. I was as well got a private beta access for a glimpse into the future of blogs search provided by Twingly. Frankly, blog search has been a daily habit for me in order to look for the new startups to be profiled on TagEdge. As a Chinese educated and a third-generation of Chinese immigrant to South-East Asian countries, I tend to use Baidu Blog Search most for of the time; otherwise Google Blog Search. In response to this Twingly, a new blog search engine in private beta phase, my first thought is that I am oblige myself to examine it whether it will become one of blogging tools to be used in the future.

Two questions while I evaluate Twingly are how it trying to differentiate itself among all the big players such as Google Blog Search, Baidu Blog Search, Technorati, etc. in the blog search sector and will this competition drive a better blog search engine for the end-users as well as the next-generation of blog search.

If you’re giving access to Twingly, you’ll notice that below the search bar, there are few search queries entered by beta users under the topic of “Hot right now.” which they claimed these are the placeholder links on Twingly. It seem that Twingly is focused on the U.S. market right now as all the search queries such as Election, Obama, Iphone are mostly related to the American. As I type few queries to Twingly, the search results showed on Twingly can be sorted out based on Twingly Rank (their internal ranking system of blog posts), date or inlinks (this might mean the inbound links from other sites). From the search results, you can further sort the results based on your preference, by time (from anytime, last hour till last month), and languages as well.

Apparently, Twingly was trying to differentiate itself not from the aspect of frequency, since beta users will not come across the blog posts they indexed at 5mins, or 10mins ago, but from the quality of all the blogs they selected to index. The posts they indexed are reasonably reliable, from wordpress.com to CNN, Blogger.com. On the Twingly search engine, any search result pages, you can sort the result page by “Any source” or “Spam-free blogs,” and either of them, the search results will be the same appeared on your computer screen. As such, I believe spam-free search will be their most important appeal to the end-users, and their vision is to surf safe while users connected to Twingly at anytime, anywhere.

The ranking of blog posts on Twingly is very straight-forward. The post that comprised of the top number of keyword you’ve entered to Twingly or with the most “Links/likes” will always go to the top, as shown in the picture below. The offering of their blog search engine want to bring values to the end-users in a dual way, i.e. the algorithm “keyword” manner and the other user-determine manner, which mean beta users can vote their favorite blog posts and these posts will eventually rise to the top of the search result pages. In designing a complete user experience, a most thoroughly discussed and commented post should stay on the top for this reason.

I think Twingly is trying to be very open, and I like their way in operating their startup in this mode. They’ve a Tech Plan, which allowed beta users to submit suggestions on what they perceived Twingly as a startup in the future. The most wanted feature at this moment is blog claiming feature, which I think some beta users desperately want to claim the ownership of their blogs on Twingly. Others are a blog profile page, a Mac version of Twingly screensaver, a real-time visualization of the blogosphere in a map.

In overall, Twingly is not very convincing in term of fast indexing, frequency when compared to Google Blog Search, or even Technorati. I think the time is still not right for them in seeking for an architectural change, i.e. spam-free blog search instead of fast indexed blog search. Safe is not fast, and either one will require the other. However, Twingly will be included in one of blogging tools in my daily life.

Microsoft has relaunched Farecast a couple of hours ago. Yesterday while I tried to access to this site, it was down. With the new logo, Farecast is apparently being built into part of the Microsoft Live Search platform.

In April this year, I did profiled a news pertaining to the purchase of Farecast by Microsoft, and even wondering whether Microsoft will going to merge the MSN Travel and Farecast, but with this relaunch site, Farecast will still remain as the same old Farecast that we know, and virtually nothing changes can be found on the landing page, not to mention the change of Farecast’s corporate logo. The only small change I find right now is from MSN Travel, where users will find airfares prediction and deals by Farecast in a small column located in the left hand side of MSN Travel.

After few days of off-blogging activities, Powerset is the next big thing that I waiting for and be the first that brought me back to blogging life. I’ve to admit here I’m a big fan of Powerset, as I did profiled several posts on this alpha search engine. If you read Powerset, it is a search engine that based on natural language understanding technology, and has taken on a new, and yet important complexion. Typing a few keywords to Powerset search engine will probably not making you understand what its core technology and how it’s differentiate itself from Google or Baidu. But you definitely aware of its organization skills, and a lot of Ajax effects on its search engine. It’s not that surprise, I read that Powerset’s interface was built by Ruby on Rails, and this programming framework has thus become a ubiquitous tool for Web development.

Although several search engines have enjoyed a significant presence on the Web, there’s still not cater to all the needs of the users. In other word, there are still many opportunities for improvement. Playing around Powerset will introduce users to a new view of what a search engine can present to the users. The launched of Powerset’s first phase product is a clear example of how a search engine can make in order to be a most appropriate tool for interacting with users on all the archive, history news. Put an example, I just input several keywords on football, and especially my beloved football club, Arsenal. The results appeared remind me how great this football club it is, albeit that I deeply sad for the fact that Arsenal did not won any trophies for the past two years. I know they’re not afford to buy a lot of superstars, but without all those big names, I still find several Arsenal success stories on Powerset’s Wikipedia article search results. If Powerset continue to improve its natural language search technology, historical news might be the mainstay of most users’ retention strategy.

This is my first thought of Powerset. Frankly, I need a second look on it. As Powerset still positioning their search strategies and framework, I’m looking forward to use it as a search engine for historical news.

If you want to know more about Powerset, there’s a video I’ve embedded in the below. Also read the key man behind this search engine, Barney Pell’s words on it.