Better Software Tools for the Humanities and the Social Sciences: a Computer Science Perspective

Better Software Tools for the Humanities and the Social Sciences: a Computer Science Perspective
Stephenson, Russell, Idaho State University, steprus2@isu.edu
Kantabutra, Vitit, Idaho State University, vkantabu@computer.org
As computer use has become more prevalent in all areas of academia, more and more scholars in the humanities and the social sciences have begun to realize the potential usefulness of software tools for their respective areas of scholarship. Many have become expert in the use of software packages ranging from database management systems and GIS to graphical and Web-authoring programs, not to mention the more common packages for word processing and the creation of presentations.

These scholars have also, however, discovered the limitations of current computer software. The problems range from lack of user friendliness to gross inefficiency and the inflexibility of database systems, to the point that important analytical discoveries which should be made are not made. We believe that these problems occur because these software packages were created to serve limited business purposes, and were co-opted for academic use simply because there were no better alternatives available.

We will discuss how to improve software for the humanities and the social sciences by designing programs based upon computer science principles that are at the same time grounded in the needs of humanities scholars, using only those software libraries, programming languages and paradigms that truly fit our purposes. In particular, we will first explain briefly how our Intentionally-Linked Entities (ILE) database system can be used in humanities research, social networks and temporal GIS as a replacement for relational database management systems. In ILE, relationships are represented using linked data structures rather than two-dimensional tables, following the long-term trend in other areas of computing towards replacing arrays by linked data structures. Relationships of user-defined complexity can be used routinely, and unlike in the relational model, such relationships are favored over the use of attributes. We will discuss why using ILE instead of other databases can lead to more accurate modeling of historical knowledge, as well as an almost complete elimination of redundancy.

We will present a new software tool we are developing that will allow humanities scholars to enter and analyze data. Humanities scholars often have documents that contain numerous pieces of interrelated information that need to be analyzed to reach an understanding of the material under study. These texts, most often word processing documents, are developed over a great deal of time from primary sources, requiring a researcher many months or years to compile. It is only after this monumental task that the scholar is finally able to start piecing together the information that is key to an understanding of the research topic. These scholars typically collect data from numerous documents containing interrelated players, places, and events. We will present a tool to help gather these disparate pieces of information into a database. The user interface for this application follows the well-known computer programming paradigm of the Integrated Development Environment (IDE), whereby a project is created by collecting relevant documents (source files, resource files, and supporting data) into a single project file defined by the user. As documents are added to the project, they can be analyzed by selecting text and creating entities and assigning relationships to those entities via an ILE database. The text and matching text in other documents may then be color-coded to make it easier to keep track of the entities that have been created. The entities/relationships will be stored in an ILE database and displayed as a tree next to the document window. We will show an early prototype of this application as well as give an overview of the fundamental components: project browser, document windows, entity/relationship trees, and the ILE database. We will also discuss applications for geographically-integrated history and its relationships to social networks and researcher-guided textual analysis.