Popular
| Summary |
|
|
In Semantic Web all data is linked, thus many users have to share common meta-data. If different users try to edit the same meta-data they have to execute transactions. The classical transactions you know from the existing data-base systems don't work because they misuse the atomicity aspect of ACID-Semantics and this decrease dramatically the transaction-throughput. Our main-product (SDMS) is a large data-grid which allows describing the data by editing the meta-data by many users. The existing data-base-systems are deficient for such complex task. Thus we developed new revolutionary concept (Multi-Isolation), which allows executing transactions by using real isolation and having constant throughput rate. What is Multi-Isolation?Multi-Isolation is a java-Framework which implements the new algorithms from our research how to isolate many variables in one atomic operation without misusing the atomicity. Multi-Isolation extends the standard isolation-process, which can isolate only one variable and adopts it. Thus you can adopt the existing multi-threaded source-code by having less overhead. The algorithm of multi-isolation can be implemented in any object-oriented languages, thus if you'd like to cooperate with us and use it for your purposes contact us! In Semantic Web we expect a large number of transactions on the same shared data, and we think the multi-isolation is the way to handle them all efficiently by having constant throughput-rate. In the image below you can see many users, who start many transactions. We expect that the 99% of all executed transactions will be executed on the meta-data (data which describes real data). Our main product SDMS must be able to handle a large number of data, thus we expecting to have a lot of meta-data. Because of collective intelligence, many users should be able to access the same meta-data concurrently and add their knowledge to the network.
Features:
Plans:
Papers: |
Happy-Guys


