How Vertical Do You Want Your Search Engine To Be?

August 19th, 2009 by Charles S. Knight
Posted in Global, Verticals | 1 Comment »

2009-08-19_1800In 2007 we saw the vertical search game play with entries like Rollyo and others that created that Google introduced their service Custom Search and Yahoo Search Builder. Those services allows you to set a list of Urls or domains to focus your searches and integrate into your site.

robi_bigBut Buscaplus allows you even more. You can define all the variables that search engines like Google, Microsoft or Yahoo use to optimize for their private tests to personalize your perfect search service. From relevance depending on word location, to relevance of links, words inside anchors, type of resources, etc…

Sites

Here in Buscaplus we offer this kind of services. We call them “Sites” and offer “Sites” and “Sites Plus”. The Sites plus allows 2,000 urls or domains. The difference is that we offer the possibility to modify all the variables that affect relevance in a search engine, giving users much more power to personalize.

Topics

This service allows for a list of tags or keywords to create a search upon. Instead of sites, this one will search the web for the topics you enter. Also allows defining the search variables that affect ranking and relevance. I set up a search account with the tags from my blog to search for those in the net.

Internet Search

But probably the most revolutionary thing is to define a search account for the web (not sites or topics limited) and adapt the configuration for our vertical purposes. Our vertical services can be totally vertical, oblique or very much horizontal. You decide on how vertical to define things in Buscaplus.

Internet users when searching have a set of preferences. This preferences go from keywords entered shown in the title of search results, images integration, results related to the query entered, etc… When Bing appeared a few weeks ago I read on blogs that some people liked it (giving reasons), others did not like it (their reasons too) and other even loving it. All those users had search preferences that were not met (or were met) by this new search engine. What we want to introduce is that the Internet community defines these preferences so user may use search account A or search account B depending on those preferences.

Examples

We introduce some examples of personalization and verticality:

  • Topic Categories: Search services targetted at content from web sites from topic categories like Arts, Internet or Computing (we have around 140 categories). Also you can define a categorized PageRank, setting relevance of links from those sites. If you build a search service about computing, you probably want to define high the relevance of links coming from computing sites. You can also define the link relevance site by site.
  • Presentation: You can develop a search theme to integrate your presentation needs. We offer consulting work for any requirement you may have about this.
  • Preferred Sites First: You can collect data from your users for their prefrence in blogs, sites, etc… and define a relevance for those in the framework.
  • Resources Preferences: Any preference related to resources like content type, language, geographic zone, etc…
  • Word Location Preferences: Preferences about relevance of words depending where they are located: Title, META, Body, Bolded content, Headers, etc…

Finally, you could define a search service horizontal with a link relevance mapping you define, preferences you gather from a set of users, etc…

We have many more options from the framework, some of them are manually operated (through editors or human feedback) which could take some time to work 100% because of the content of pages for those modules. Other computer-related modules are operative and give you personalization right after you create a basic free search account.

Source: Buscaplus

Conversation search engine uberVU

August 19th, 2009 by Charles S. Knight
Posted in Uncategorized | No Comments »

2009-08-19_1729uberVU is an easy way to find and follow conversations, even if they take place across multiple sites and services.

The interesting stories you read on the Web don’t show you the whole picture. The conversation around them is missing. What people say about those stories is a great part of the picture.

You might upload a video on YouTube that gets embedded in a blog post. That post gets comments and it gets posted on Twitter, where it also gets some replies. The Twitter post gets to FriendFeed where the conversation continues.

tour_findAll of this is part of a single conversation, but you can’t see it because it’s trapped inside different services. That’s where uberVU comes in.
Find conversations

You can start by entering a keyword or the URL of a conversation that you are intersted in. When you search for a keyword we don’t look for just for blog posts or twitter messages, we scan entire conversation threads to show you things in context.

So if you search for “keyword 1″ we may find a conversation that began with a blog post without any connection with “keyword 1″ but where people are discussing it in the coments or on FriendFeed.

2009-08-19_1722

Because uberVU gets reactions to stories from blogs, Twitter, Digg, FriendFeed, Disqus, YouTube and many more, we figure out the relationships between all those reactions. This allows us to provide you with more search depth than ever before.
Explore

Once you opened a thread you find the whole conversation in one single place.

Oh, and did we mention that we also keep you up to date on how the conversation evolves? All you have to do is check uberVU for updates

No more searching a dozen services for mentions of a story, no more keeping track of all those people that commented on the story all over the web.

Source: uberVU

S. O. S. Search Our Ships with MarineTraffic.com

August 19th, 2009 by Charles S. Knight
Posted in Verticals | 1 Comment »

shipping_smallThis web site, MarineTraffic.com, is part of an academic, open, community-based project. It is dedicated in collecting and presenting data which are exploited in research areas, such as:
m- Study of marine telecommunications in respect of efficiency and propagation parameters
- Simulation of vessel movements in order to contribute to the safety of navigation and to cope with critical incidents
- Interactive information systems design
- Design of databases providing real-time information
- Statistical processing of ports traffic with applications in operational research
- Design of models for the spotting of the origin of a pollution
- Design of efficient algorithms for sea path evaluation and for determining the estimated time of ship arrivals
- Correlation of the collected information with weather data
- Cooperation with Institutes dedicated in the protection of the environment

It provides free real-time information to the public, about ship movements and ports, mainly across the coast-lines of Europe and N.America. The project is currently hosted by the Department of Product and Systems Design Enginnering, University of the Aegean, Greece. The initial data collection is based on the Automatic Identification System (AIS). We are constantly looking for partners to take part in the community. They will have to install an AIS receiver and share the data of their area with us, in order to cover more areas and ports around the world.

Source: MarineTraffic.com

Search engine Feedeep in under 5 minutes

August 19th, 2009 by Charles S. Knight
Posted in Uncategorized | No Comments »

index

Click here to visit search engine Feedeep

Directionlessgov.com – Google vs. Direct.gov.uk

August 19th, 2009 by Charles S. Knight
Posted in Uncategorized | No Comments »

2009-08-18_1749Directionlessgov.com searches more than 6 million pages of UK government info, instantly using direct.gov.uk and Google search.