Wo andere Suchmaschinen versuchen den Kontext des Inhalts aus Keyword-Assoziationen zu berechnen, soll NetBase einen Schritt weiter gehen und auf Basis einer Analyse ganzer Sätze die Assoziationen von Sätzen und Bedeutungen zueinander automatisch erkennen. Somit soll Content Intelligence auch ganz ohne vorherige Wissens-Modellierung oder Lexika auskommen können.

Damit folgt NetBase dem Trend weg von der Suchmaschine die Suchergebnis-Listen ausspuckt, hin zu Frage-Antwort-Maschine a la WolframAlpha. Netbase wird allerdings vorraussichtlich keine eigene Suchmaschine anbieten (abgesehen vom Demonstrator), sondern auf ein Lizenz-Modell für Unternehmen setzen – lt. TechCrunch soll die Lizenz bei 100.000$ anfangen. Diese Kröte sollen dann Unternehmen schlucken, die ansonsten große Aufwände bei der Modellierung von Taxonomien/Lexika o.ä. in Kauf nehmen müssten.

Doch adressiert NetBase damit tatsächlich ein Problem oder handelt es sich lediglich um ein Sahnehäubchen?

Artikel in der Technology Review: Smarte Gesundheitsdatenbank

Breakfast Links 08.01.2009

* TrustYou – make a good choice

I just stumbled upon a press release from TrustYou, a search engine for hotel and restaurant recommendations. TrustYou calls itself a semantic meta search engine. Meta because it makes use of various existing hotel recommendation platforms like HRS, hotel.de and more. Semantic because of its pre-processing methods for heterogeneous text data about hotels/restaurant and user comments. Yeah, finally a new alternaitve search engine from germany, chapeau!

See also this post: TrustYou durchsucht Hotel-Bewertungen (german)

* Google Semantics Surfacing

Stephen Arnold’s response to the article “Did Google Just Expose Semantic Data in Search Results?” from readwriteweb.com.

I wasn’t able to write any Breakfast Links since monday due to projects (like the upcoming relaunch of this blog), birthday parties and efforts for christmas presents (big, big family)! Finally I turned on my news reader today…

* Amazon Launches Public Data Sets To Ease Research

Amazon (AWS) provides public data sets to demonstrate use cases for their (cloud) services. The article on TechCrunch also links to Swivel which is a repository for open data sets. Unfortunately you won’t find any textual data but a lot of data like U.S. Census, historical oil prices, world photovoltaic production and more.

* New Open Source Search Vendor

An interview of Hans-Christian Brockmann (Brox IT-Solutions GmbH) about Brox Open Source strategy in the field of enterprise search. Really interesting is Brockmann’s number of 180.000 lucene projects (more or less productive) out there! Brox already came up with their Eclipse project SMILA, which is going to be a base architecture to better integrate search and text mining into common systems. Nice to hear from those kind of projects in my business area!

* Z! – Episode 134: Qimaya – Anwendung neuronaler Netze (german)

Interesting interview with Roy Uhlmann (Qimaya aka former Queap) about neuronal networks in their search engine Qimaya. If you are interested in testing Qimaya you can request access to the demo – next invitation and access delivery is even today! As I can see, Qimaya is currently a german language only project!

Breakfast Links 03.11.2008

* Exalead-Labs: voxalead

The parisian search engine company Exalead released their video search engine voxalead. It combines Exalead’s indexing (Entity Recognition) and search technology with third partie’s Voice recognition (LIMSI). Currently they indexed mostly french and us channels and a small amount of german channels. Beside the search through the voice-to-text extracted contents you can also view the original extracted text which shows off the quality of voxalead!

* Metager at Alt Search Engines

Metager now has its place at AltSearchEngines.com in english and german!

* Mahout: Twenty Newsgroups Classification

This weekend I stumbled upon a nice Mahout Classification example. It is all about creating a classification model based on twenty clusters of newsgroup messages. All you need is Mahout and Hadoop to get the example running in approx. 20 minutes.

Breakfast Links 15.10.2008

* Yotify: A Social Search Engine

Stephen Arnold reviews the Social Search Engine Yotify regarding to the article on the Technology Review magazine.

Breakfast Links 08.10.2008

* Ask.com with new version of search

Ask.com with changes in terms of search relevance, search speed and question answering. But why am I getting 10 to 15 sponsored links in some result sets?

* Autonomy and OpenText: Equally Good Bellwethers

Stephen Arnold compares Autonomy and OpenText and gives an outlook into the future of them and enterprise search in general.

Breakfast Links 07.10.2008

* Welcome to the Data Cloud

* VN (vietnamese) search engine released

* Three outstanding graduate theses – with a really interesting language-based approach in categorization -  on IR Touthgs

Breakfast Links 01.10.2008

* Schleswig-Holstein communities saying no to Google Street View and planning legal actions to protect themselves from Google

* Whole Travel beta released, a travel search engine with the possibility to rank results by sustainability

Today I read some articles about Simplexo, a new open source enterprise search application provided by a privately held company from the UK (Simplexo Ltd.). Since I’m already experienced with open and commercial enterprise search solutions I was asking some questions in terms of features and open source to Simplexo’s website (please keep in mind that this is not a test in practice).

What are the system Requirements, Hard- and Software?
Simplexo is a Microsoft only product, it needs Windows XP, 2000, 2003 or Vista, the IIS server, .NET 2 Framework and finally a relational database like MySQL or MS SQL.

Which kind of files are getting indexed?
Simplexo already comes with parsers for structured and unstructured documents. It can handle documents like MS/Open Office, PDF, PS, Webpages, Flat-text, Images (I guess metadata) and E-Mails (?). Beside those files from you’re local filesystem or network drive, Simplexo can also index information from databases, business applications like salesforce or sap and online sources such Google, Reuters and arbitrary feeds.

Is it possible to integrate modules for content processing?
I couldn’t find any information on how to extend Simplexo on their website and I’m still missing information on mailing lists, developer forums etc.

Which search functions are available?
Simplexo can handle ordinary term, phrase and boolean queries. Beside that Simplexo also offers Synonym support, date range, fuzzy and semantic queries (detail information missing).

Is it possible to integrate Simplexo in existing applications?
Yes, Simplexo comes with a SOAP interface and already provides an interface to search through existing Microsoft and Autonomy indexes.

What about scalability?
There is a page about scalability at Simplexo’s website but it only tells you about 600k simultaneous users, less search time than Google and a much smaller index size than on MS databases. I’m missing information about how the Scalability is done technically, is it possible to use more hardware?

So, what about a test in practice? Lately there is no online demonstration of Simplexo yet and since I’m running a Linux desktop I can’t perform the installation of *.exes – Simplexo is a pure Microsoft compatible product.

Original press article on CBR: Simplexo launches open source enterprise search platform

Breakfast Links 11.09.2008

New Video Search Engine: VideoSurf (I’m still waiting for my invitation, can anybody help?)
http://searchengineland.com/080910-050000.php

Cascading tries to simplify building Hadoop MapReduce Applications
http://www.theserverside.com/news/thread.tss?thread_id=50629

Free Application Servers tested: Tomcat vs. Geronimo vs. JBoss vs. GlassFish (german)
http://www.computerwoche.de/knowledge_center/software_infrastruktur/1873146/