Details of design

ScienSe is a system developed under the client-server model. On the client's side there is a web interface based on HTML, CSS and Javascript which allows to manage the information stored in the system. HTML is used in order to organize the information of the menus, forms and presentation tables. CSS is in charge of the graphical presentation to the user and Javascript is used to validate the input data and for present certain elements make easier the interaction between the user and the application. This validation of the data is made only for improving the user interface, since by security reasons within the components that manipulate the information those validations are made again. All the interface is built-in on the servlets, thus, in order to modify the interface, it is necessary to know this technology.

On the server's side the system consists of 3 main components:

The first of them is a relational database that stores information about the research publications, the database, and all the statements that manipulate the elements contained in the database were developed using SQL92 so as to allow an independence from the database manager system. Within the source code of the application are the tables and the entity-relation diagrams of the database.

The second component is a Web application on the server's side which allows to manipulate the information contained on the database. This application was developed using Java program language and technologies as J2EE, Java Servlets and Java Server Pages. The connection with the database is made by means of JDBC.

The third component is a program that generates an index of the Web sites given by the user, there it looks for the information of scientific publication and stores that information so as to add it to the database, later, after being checked by the user. If it finds files in formats as HTML, PDF, PS, DOC, among others, it uses a filter in order to turn them into ASCII text and then apply the identification of data algorithm. All the generated information is stored in a temporary file, which when the indexation process has finished, is read by the web application in the check publication process. wget is used for download the documents in order to analyze them. It was developed using ANSI/ISO C/C++ and uses the system POSIX's interface libraries.