, , , ,

It is well known that we live in what can be called the “Information Age” and that we are constantly under the influence of different ideas that try to shape our minds. This fact is of primary importance in Indonesia [1], which, in 2014, was the eighth-highest country in the world concerning the number of Internet users. This is an amazing result especially if we recall that Indonesia moved from a repressive to a democratic regime just in 1998. The Internet is seen as a powerful means to promote democracy, being it the most open and free form of communication available nowadays. However, young people need to learn how to effectively interact with all the available media so as to develop a sharp and analytical mind and enhance democracy in Indonesia. To this aim, digital media literacy (such as blogs and social media) is perceived as an important life skill to understand and support democratic institutions and practices.

Among all the media we are surrounded by, audiovisual material will probably represent (if not yet) the first source of information. Such a means is of primary importance in Digital Humanities since a single document can provide a lot of information regarding not only languages and the content itself but also the context it is linked to, such as the place where it took place and the people involved in. Scholars in Digital Humanities are still starting to use such means due to several issues. One of them is the ability to efficiently search among the huge amount of audiovisual data present in the Internet. An approach which tries to provide an easy access to such libraries is given by the AXES (Access for Audiovisual Archives) RESEARCH system [2]. It has been built to meet the needs of humanist scholars and journalists, providing an advanced retrieval system based on different technologies: from face and place recognition, to similarity searching and speech recognition. Thanks to its backbone, users can find information they are interested in using several approaches, namely: text/spoken words search, visual search, and similarity search. The first approach is based on the creation of transcripts of the original files using automatic speech recognition. The searching operation is then performed between all the stored metadata and the key words the user types. The visual search approach retrieves objects in the archive that are related to a text that is typed by the user. This text is used to find visual objects in search engines, such as Google Image Search, to retrieve an image representation of what the user is looking for and eventually the system shows the user objects from its storage which are similar to the results given by the search engines. The similarity search approach, instead, does not require the user to enter keywords but just to provide one or more images. The retrieval process is similar to the previous one.

The ability of extracting transcripts from audiovisual materials is of fundamental importance in several situations in which people are directly or indirectly involved. For instance, in California, important legislative decisions occur in committee hearing, which are video recorded and then made publicly available. However, it takes a lot of effort to find relevant information in hours and hours of videos. With the aim of giving access to these transcripts to all the interested users, the Digital Democracy (DD) platform was created at California Polytechnic State University [3]. Data, such as video archives and the nature of legislative sessions, is stored in a database and, automatically, accurate information regarding a specific file is then extracted and made available to the final user in a web portal. To obtain some feedback during the development of the project, several categories of possible users (from individuals interested in legislative affairs to legislators) were asked their opinions and all of them were enthusiastic about the results, underlining the primary role it might take in the future of California.

Thus, multimedia represents a rich source of information and it is going to become the primary source for future Digital Humanities scholars. As it has been described above, there are still some difficulties in extracting relevant data from these information containers. A tool as AXES RESEARCH implements some features that enable acquisition and extraction but only a few of such systems are currently available. People are really excited by these projects and find them very useful. Plus, the abilities of such systems may well increase as they becomes more and more common. An example has been given for the extraction of important information from legislative discourses in California via the Digital Democracy platform. People can now either have access to a given transcript or directly access automatically-selected relevant information regarding that hearing. Moreover, the possibilities that multimedia offers do not stop in implementation technologies but are actually shaping tomorrow’s human societies. One significant example is given by Indonesia, where several studies are going on in order to assess the ability of young people to effectively use digital media as means for better supporting democratic institutions.

The application of digital methods for analysis of audiovisual data is still moving its first steps but it represents a key concept in Digital Humanities due to the large amount of information that is contained in them and we will probably hear a lot about them in the next few years!


[1] (2015) Fiona Suwana: Digital Media Literacy in Indonesian Youth: Building Sustainable Democratic Institutions and Practices

[2] (2015) Martijn Kleppe, Max Kemman, Peggy Van der Kreeft, Kay Macquarrie, Kevin McGuinness: AXES: Researching & Accessing Videos Through Multimodal Analyses

[3] (2015) Sam Blakeslee, Alex Dekhtyar, Foaad Khosmood, Franz Kurfess, Toshihiro Kuboi, Hans Poshcman, Giovanni Prinzivalli, Christine Roberston, Skylar Durst: Digital Democracy Project: Making Government More Transparent one Video at a Time