--- a/paper/abstract_eng.tex	Sat Feb 01 16:15:35 2014 +0900
+++ b/paper/abstract_eng.tex	Sat Feb 01 17:08:52 2014 +0900
@@ -3,17 +3,21 @@
 It has caused the webserver is down.
 Therefore, scalability is important software factor today.
 Scalability in distributed system is able to increase performance linearly when just added new node to system.
-In order to make provide scalability, we considered design of architecture.
+In order to make provide scalability to Web services, database must have scalability.
 
-We are developing a database Jungle. 
+We are designing and developing a database Jungle for study of scalable database. 
 It is use non-destructive tree structure.
 Non-destructive tree structure is not the destruction of data.
 Editing of data is done creating by new tree.
+
 Jungle was designed as a distributed database.
 But data distribution and persistent has not yet been implemented in the Jungle.
 
-In this paper, we develop distributed database on jungle for pursuit architecture with scalability.
+In this paper, we develope persistent and distributed database on jungle.
 Distributed data on Jungle is developing using parallel distributed framework Alice.
-As a result, we confirmed that data is distributed  between the server node.
+We implemented distributed on Jungle and confirmed with our cluster system that data is distributed between the server node.
+Also, we developed simple bulletinboard system with Jungle and key-value store database Cassandra.
+We compared Jungle and Cassandra using simple bulletinboard.
+As a result, we got better performance then Cassandra.
 
 \end{abstract_eng}