Archive for the ‘Search’ Category

Involver readies personalized event search

March 5, 2007

I have been watching out for the launch of Involver for the past few months. Involver’s event search engine has recently entered a private beta, so I thought it is a good time to give it a try. Involver basically tries to marry personalization with event search. It will be looking at your interests, your friends(and their interests) to deliver personalized events results to you. In you can say – for event search. For the past week that I have been trying Involver on-and-off, I would say that the implementation and integration is pretty good. Finding similar events, direct link to buy tickets from Ticketmaster, adding to calendar and watchlist, commenting on events, and adding friends to your network is as effortless as it can be. I think it would be early to give any decision on the personalization part since you need far more data and big enough friend-network to get meaningful results.

From the integration standpoint, Involver is another of the recent startups that has integrated 30Boxes for the calendar side of things. With Kiko long gone and Trumba changing directions, in the short history of web-based calendars I have seen 30Boxes going many more places than Google Calendar which hasn’t seen as much traction with the developer community. I think not many people saw this coming.

Coming back to event search, I think it is another of those embattled areas which has too many startups chasing not that big amount of money. Zvents, BusyTonight, Eventful, Meetup, Upcoming, HeyletsGo, Attendio, Minglenow…….. are just some of them. Some of these funded while others in process but none of them in sight of a MySpace+- exit even though the idea has been hanging on for as long. I would like to see how Involver defines its direction.



TheFind now finds similar products

March 1, 2007

 TheFind product search engine got a major feature boost yesterday when they introduced the ability to search similar items. After playing around and trying to find similar items I am impressed with the release. The similar results I got during my testing were similar in major feature aspects yet different in their own unique way, which quickly lead to number of second options that I in fact liked. To power this feature TheFind has put its crawling and indexing effort to the best use. TheFind already indexes 150 million products from 500,000 stores to gather information including product titles, descriptions, prices, other merchandising that the stores contain. TheFind engineers have applied this vast set of data to their  advantage to build an algorithm that accounts for specific attributes for each product and compares to one another to determine if there is a strong relation between the items.  Based on a proprietary matrix products are then ranked by the strongest correlations of relation and relevance. had been operating in this area for past few month. Even though they have done an amazing job at building their core product, I have never really been impressed by their heavy focus on using digital signature to find similar items.

Rethinking the feature release, TheFind has done a good job at the initial launch of similarity engine. Still I think either the results need to be further fine tuned to enable faster drill down to the right product.


Quintura upgrades; Now searches videos, Amazon

February 26, 2007

Quintura, the visual search engine whose funding news had spread around from last year, has released a new version of their See & Find™ concept based platform. With the launch of this new release, Quintura has introduced the ability to search videos and products at Amazon. Quintura is currently using the Yahoo Video search to power its video search, and obviously hitting the Amazon API for other addition.

Quintura has also started displaying graphical images (favorite icons) next to tags (search terms) in the tag cloud. The icon is associated with a URL that corresponds to a search term. Clicking the icon from the tag cloud takes you right to the site, which gives the Google like “I’am Feeling Lucky” type convenience as you move from one tag to another. Also with the new release, Quintura has moved the results on to the right side of the page, which I think is a much better placement as compared to having the results beneath the tag cloud.

If you haven’t given Quintura a try, you should consider that doing soon since it brings out so many tag based relationships for your search keywords that you will be amazed at results you typically can miss out on Google.

Quintura – Visual search engine startup gets funding
Quintura confirms funding from early Skype investor
Quintura provides safer search for kids

Mojopages local search goes alpha

February 21, 2007

Mojopages, the local search engine that I had written about in the past, finally launched alpha version of the site over the past weekend. From the initial view, the site looks pretty comprehensive that adds a very strong social networking angle to plain old local search.

Backed by commercial local listings data, Mojopages enables users to rate and write reviews of businesses. Number of interaction points build into the platform gives the site a dynamic feel. Mojopages enables uploading of photos, videos, sending IMs, retrieve coupons, and many more related options. You can easily create social networking site worthy profile, add people to your network, join discussions in the community via QAs and SmallTalk. Best part I like about Mojopages is the emphasis on tracking user activity and contributions to the community. Typical problem reading reviews on a regular site is that you are not really sure whether the review is genuine and trustworthy. At Mojopages you can quickly view all the reviews and ratings given by a particular user besides their activity in community QAs, and Small Talk.

Having said that, I think trying to marry off local search with social networking is tough task. Quite a few have tried the marriage, some have changed directions after not being able to make headway(read Judysbook), some are going down slowly(read CitySearch), and some gaining ground albeit slowly(read Yelp). Would be interesting to see how Mojo plays the game differently and manages to win.


AdaptiveBlue’s semantic plugin gets funded

February 20, 2007

AdaptiveBlue, startup behind the cool Firefox extension BlueOrganizer that launched last year in September at DEMOfall, today announced that it has closed Series A round of funding. The financing was lead by Union Square Ventures. Financial terms of the funding were undisclosed.

Yesterday was the first time I gave a real try to BlueOrganizer after a session with Alex Iskold who is the founder of AdaptiveBlue. I was amazed at the BlueOrganizer ‘s semantic relevancy engine built out into the extension that lets you get done with lot of things faster. Consider you are on, checking out movie reviews to add your Netflix queue. Regular tedious process is to search in IMDB, copy the name of the movie you like, and than paste/search the information at Netflix once more. BlueOrganizer gives you a simpler option for this. Just go to the BlueOrganizer right click context menu, and the app smartly recognizes the site you are on and gives you an option to add the movie to your Netflix queue right away. Such contextual hookups from  BlueOrganizer come up for each of the sites you visit everyday spanning hundreds of small to big categories. In the background BlueOrganizer works with as many APIs and search URL manipulations possible to drive down on the number of clicks needed for everyday tasks.

Best part about AdaptiveBlue is its focus on maintaining privacy around user data and still be able to provide personalized results unlike Google which has its own set of issues. The privacy part comes into play when you want  to quickly get your list of personalized sites for the contextual linking in BlueOrganizer without storing your browsing history on a 3rd part server. All you need to do is, hit the “Personalize” button in the BlueOrganizer’s settings, and it goes through your browsing history, looks at the frequency of the sites you typically visit, runs some more analysis, and automatically loads your personalized site list. All the while, none of your browsing history related data ever gets moved over to AdaptiveBlue servers. Seems like a simple but smart solution to me.

There is lot more to BlueOrganizer which you can discover once to start playing with it. Personally I think I would be using frequently from now on.



Systemone’s Similarity Engine goes alpha

February 19, 2007

Semantic Web is the buzzword these days. TextDigger, Radar Networks, Boorah, PowerSet, Adaptive Blue are just some of the startups working on semantic web variations. Another semantic web contender Systemone today launched their first consumer facing Similarity Engine named Infolust. The initial launch is very basic. Right now all you can do is feed in the URL of a page for which you want more details, and Infolust will quickly analyze to come back with 10 related pages from Wikipedia. As I mentioned before the this is an alpha product so it is not yet ready for prime time. From the few search queries I ran, most of the results were way off track. So lot of work needs to done in that aspect, which even Systemone openly admits. Besides this, Systemone will be adding more content sources beyond the Wikipedia domain. Even if Systemone can get Wikipedia part working right, it will be big improvement over the standard Wikipedia search which is no one’s favorite.

Another similar product that I came across few days back is Watson from Intellext whose search results are completely and totally lost.


Google Search history can get you in a spot

February 2, 2007

I had been feeling uneasy about the Google Search History option for sometime. Today found the perfect reason to get rid of the option from my Google account. Karen, Vice President of Marketing at Leaptag which is a personalized web content discovery tool, today described on her blog how her search history ended up being viewable by someone else. Not a good feeling when your past 2000 searches are available for someone to get insights on you. Right now, my search history is gone for good(I hope that is what happened when I hit the delete button).

If Google listens, here are few things Google can do for us – Let us view and manage logins for all the machines we are logged on at any moment of time. There should be another key/password to get entry to such a login manager. Provide a button for Google Toolbar that shows the logins for each of those machines in a drop-down. Also by default, “Remember me on this computer” option should be unchecked on all Google Services.

Any more ideas?

Attendio tries event search(seriously)

February 2, 2007

Ah! Local search, location based search, and all those life and time savers. Attendio too wants to join that league with its local events search engine. You can search for – Music, family, Perf Arts, Fashion, Film, Technology, Public Affairs – all in one place(and only for Bay Area currently). Users get to add events, reviews and recommendations while Attendio provides VIP “Recommenders” for insider’s view on all the “hot” events, which sounds more like MySpace beginnings. They also number of integrations to deliver event information to calendars(GCal, iCal, …..), and email. On mobile you text Attendio your interests and zip code, and they come back with the highest ranked event matching your interest.

Anyway, we will soon know whether Attendio is just another search engine to forget or can rise above that. What do we already have in the market? – Zvents, Eventful, Upcoming, to name a few.



Farecast goes for Series C

January 30, 2007

Farecast, one of the best and my favorite search travel search engine, has stomped home with another round of funding. According to PEHub, Farecast raised $12.1 million in the Series C round of funding led by Sutter Hill Ventures, and joined by existing investors Madrona Venture Group and Greylock Partners. This takes the total funding for Farecast to $20.6 million from its 3 rounds. 

Farecast searches seem to much more effective compared to its competitors ’cause of the ticket price guidance it provides based on historical price data analysis. This feature combined with delivery of persistent search results for ticket prices via RSS can save you time and precious dollars.

Kayak is another startup giving price guidance but without RSS feeds.  


Textdigger to launch semantic search engine

January 30, 2007

TextDigger will be launching its semantic search engine at DEMO today. TextDigger is not another one in the series of startups that are launched every now and then claiming Google is pretty bad at what it does and they will make sure that the injustice is stopped. TextDigger does something important which adds gives an interesting dimension to search space- Search engine being able to find similar words in context to the keywords your searched for. And these are not the keywords taken out of thesaurus we are talking about. Take the case when you are search for “hotel with a view”. The keyword has various dimensions to it – View as in “point of view”, View as used in engineering drawing, and more. However what a user is looking for is pointers in the direction of “vista”, “panorama”, “landscape”……, which a user might not recall and needs at that point of time.

When you search at TextDigger for any string, it will come up with all regular search results, and on the top of the page gives you the semantic results that can help you refining your search results. TextDigger will be initially opening up the search engine to a select group of beta testers. Semantic part of the search results will operate in a social search fashion with the users having the ability to edit and add to the semantic search index. In case you are wondering about the how they going about building their search and semantic index – TextDigger has built the semantic search engine ground up, while partnering with Gigablast to deliver the regular search results.

Coming to the short history of TextDigger – based out of San Jose, CA, TextDigger was officially started off early last year by 3 lead CNET engineers. Till date TextDigger has received funding from CNET and several angel investors but haven’t done a VC round of funding. Currently TextDigger has 6 people onboard, with varied background in CS, linguistics, and philosophy which gives it a balanced outlook at how users perceive search to be.

Powerset, and Phrasetrain have been working similar problem with slightly different focus in each case. TextDigger team works a lot around linguistics while other might try to solve them the perspective of grammar.