| 研究実績の概要 |
In the third (last) year, we have done the following 3 topics: 1) More General Evaluation of service discovery performance on Map-Reduce Global Social Service Network (GSSN) 2) Another GSSN Construction using Word Embedding technique using Tensorflow 3) Broader application of Big Data infrastructure.
As evaluation of this research through entire periods of 3 years, the objective and plan of this research is to develop distributed algorithm and system to discover services on GSSN and its evaluation and application. It can be summarized as below.
First, a novel algorithm, called Map-Reduce Global Social Service Network (MR-GSSN), to generate large service network, has been developed on Hadoop cluster with 18 nodes. We evaluated service discovery performance based on MR-GSSN, and it shows almost same result as that of GSSN. As computation performance had been increased by 30 times comparing to non-distributed environment.
Second, in this research, we proposed Linked Social Service Network (LSSN) with multiple feature attributes based service discovery for Big Data Analytics. It is a combination of two advantages, which are precision and sociability of web services.
Third, another new GSSN has been generated using Word Embedding in Tensorflow with data from social media Web application API. And service similarity calculation using a new ontology generation method has been developed. It can be applied to calculation of service link quality with higher precision. Finally, we developed a new algorithm to calculate several tasks on Big Data infrastructure.
|
