Project Goals and the Creation of Symbiota2

Project Goals and the Creation of Symbiota2

We will transform Symbiota into an improved software platform: Symbiota2. The features requested by the Symbiota community via the forum and are distilled into the following six goals for this development program.

  • Goal 1.  Provide RESTful (representational state transfer) web services – There is growing demand to use Symbiota in a scientific workflow with other applications, for example as a data provider to the Interior Collections Management System that catalogues and manages all the collections owned by the U.S. Department of the Interior, or as a step in morphological, geographic, and ecological analyses. Symbiota2 will have an expanded API framework and offer a variety of RESTful web services that will enable external applications to efficiently pull data into workflows, export data using specific filters and in specific formats, and import data into the Symbiota2 portal.
  • Goal 2, Plugin design and deployment – The project team will refactor Symbiota into plugin modules that have cohesive functionality,g., a plugin to upload scientific names from a file. Refactoring Symbiota’s functionality into plugins will create an extensible framework for development. Users will be able to customize Symbiota2 by selecting only desired plugins and creating new plugins as needed. Refactoring will also allow Symbiota to be transformed into Symbiota2 in stages rather than requiring all of the pieces to be in place before Symbiota2 is released.
  • Goal 3.  Accommodate other Database Management Systems – As biodiversity data collections grow in size, new kinds of data storage and data analyses need to be supported. Symbiota2 will offer a variety of database management system (DBMS) options to store biodiversity data. Symbiota’s data is currently stored in a relational database comprising 143 tables managed by a MySQL DBMS. Database operations are spread throughout the code and are specific to MySQL. The project team will create a database abstraction layer and object relational mappings for MySQL, Postgres, SQLite, and Google’s Cloud Spanner [16]. This will result in a database abstraction that models the data as a collection of objects and maps each object to a database table. The abstraction layer ensures that all data manipulation and querying is performed by the mapping, encapsulates the DBMS operations in a single place and, importantly, enables other DBMSs to be used by changing the mapping.
  • Gole 4. Increase Data Utility – Symbiota2 will help scientific researchers analyze and make use of biodiversity data by providing new ways to view the data and creating a framework for running analysis tools. A key part of Symbiota2 will be the ability to develop and integrate tools for analyzing data. Symbiota2 will have an external API (Goal 1) so that tools can pull data, and will also have a plugin architecture (Goal 2) so that tools can be closely integrated. To further increase data utility, we plan to export a Symbiota2 knowledge graph [30]. A knowledge graph is a semantically-rich collection of linked data. Ontologies help define and describe the data and create opportunities for semantically linking the data to other graphs. Symbiota2 will leverage taxon ontologies [59][67] to export a biodiversity knowledge graph [46][66].
  • Goal 5. Separate form from function – Symbiota’s Graphical User Interface (GUI) is programmed into many PHP files that also manipulate the data and perform other functions. Many, but not all, style decisions in the GUI are controlled by a collection of Cascading Stylesheets (CSS) rules. By cleanly separating the GUI from the core functionality the code can be reduced in size and complexity, thereby becoming more maintainable and extensible. Model, view, controller (MVC) frameworks (e.g., Ruby on Rails [54]) are often used to create such a separation, but Symbiota was developed before their value was recognized [20]. Adding an MVC framework to Symbiota will also promote Content Management System (CMS) integration. Customized content has often been challenging to add for Symbiota portal managers with little experience in web development. Many managers have taken the approach of creating a separate site to document their portal, explain workflows for data management, and provide background for their enterprise. A CMS, such as WordPress, makes it relatively easy for non-experts to install, customize, and maintain a web site. By integrating with a CMS, Symbiota2 will gain many CMS features, such as user control over the look-and-feel, support for internationalization, automated update, configurable installation, advanced security (e.g., Wordfence), responsive layout, WYSWIG editing, and social media syndication.
  • Goal 6. Data Collection Enhancements – Users want the ability to incorporate metadata (e.g., camera EXIF data for images) and new kinds of data into Symbiota, such as monitoring and survey data. In addition, they have asked for the ability to create new data entry/editing forms and data cleaning tools. Symbiota2 will be designed for such extensibility, allowing it to easily accommodate new data types and improve data management. A recurring problem in data collection is that museums in developing countries and remote areas of North America have only intermittent Internet access. To help such users, Symbiota2 will have an offline mode. Data collected offline will be synchronized with a live portal when reconnected to the Internet.