Find solution
How to handle database connections?

What is the best approach for building a kind of a semantic/ontologies database? What are the best tools/products/algorithms?

  • I'm searching for a way to build a recommendation system which can handle different types of contents like Locations/Places, Products, Movies, Books and so on. When a User enter a Book/Location/Movie Name the database should be search if it's already there (or look with an API somewhere else) and show some autosuggestions for that to avoid duplicated content in the database. After that the user should have the possibilty to rate this thing and other users too. It's a bit like Amen but with more content and their relating connections.

  • Answer:

    There's a bunch of good work and smart people out there for you to reference and use. For semantics / ontologies, http://opensemanticframework.org/ is a good start. For recommendations, Zemanta has been at this awhile, and has made code available on GitHub. https://github.com/Zemanta/simple-recommendation-engine dmoz has open directories of books, locations, movies. http://www.dmoz.org/

Steve McNally at Quora Visit the source

Was this solution helpful to you?

Other answers

You can start with any of existing approaches for Triple and other graph-based databases. There are lots of them on the market, here is a list of a few I've been looking for back in December'11: AllegroGraph | Linux/Mac OS X 4store | Linux Fuseki | Linux Joseki | Java-based | Linux/Windows/Mac OS X Sesame | Java-based Stardog  | Java-based | Commercial Talis | SaaS | Partially Commercial Virtuoso | Commercial Microsoft Zentity | for Research purposes | Specific for Scientific Repositories IntelliDimension | Semantic Platform | Commercial If you want to build your own and it is a web-service, you'd better to start with a one that is originally a scalable thing. So, you shouldn't choose Semantic Platform by IntelliDimension or Zentity by MSR. You'd better to use a system that has SPARQL support; custom query language isn't bad to have but you'd be locked to a specific vendor in that case. will be really useful to run quite sophisticated queries against your data, including those things you mentioned in the quesiton. If you write a .NET-based service, for an API layer you can use DotNetRdf, as other variations (LinqToRdf and SemWeb) are more or less deprecated. DotNetRdf is able to work with a lot of different technologies including a few of those mentioned above; it is also able to use SQL Azure to create a Triple Storage in the cloud. Keep in mind that you should store BLOB data in BLOB storage, not in database. In SQL Azure max size of a db is 150GB, so I'm not sure that it is a good choice for a really big, scalable service. If you want to have a local service, you can start with SQL Server 2012 which has its free version and Semantic Search included. That's we do at . We've created a custom Triple Storage on-top of SQL Server, added FileTable feature to store binary content, added Full-Text & Semantic Search to the system and are working on joining SPARQL query engine with FTS/Semantic Search all together. Notice that metadata is done by SPARQL; FTS/Semantic Search is used only for contents of binary files (docs, etc.).

Daniel Kornev

Related Q & A:

Just Added Q & A:

Find solution

For every problem there is a solution! Proved by Solucija.

  • Got an issue and looking for advice?

  • Ask Solucija to search every corner of the Web for help.

  • Get workable solutions and helpful tips in a moment.

Just ask Solucija about an issue you face and immediately get a list of ready solutions, answers and tips from other Internet users. We always provide the most suitable and complete answer to your question at the top, along with a few good alternatives below.