Content

Hide video with translation into sign language
Switch to fixed-width version
Česká verze této stránky English

Paper details

An interdisciplinary approach to alternative representations for images - Splendiani, Bruno


Bruno Splendiani, Mireia Ribera, Roberto García, Marina Salse
Library and Information Science Department, University of Barcelona, Barcelona, Spain Computer Science and Engineering Department, University of Lleida, Lleida, Spain



Abstract

1 The R & D or application idea

A key principle of accessible design is the use of appropriate text alternatives to non-text content in web pages and digital documents. A text alternative allows the images in a web site to be readable by screen readers or other assistive technologies, making them accessible by blind users or users with other disabilities. Although the need of text alternatives is recommended by the number 1 guideline of WCAG version 1.0 and 2.0, currently the implementation of alternative text to images on the web is far from being totally correct and lots of images have missing, incorrect, or poor alternative text (Petrie et al. 2005). The reason could be that the normative only offers general recommendations regarding “how” to create the alternative text description for images and currently there are not commonly agreed guidelines on how to describe the image content nor standards defining the process of the image description. As an answer to this limitation, several attempts have been made to formalise the description of images by different institutions, being the most relevant the set of guidelines for describing images proposed by the NCAM (NCAM 2009) and the DIAGRAM Center (DIAGRAM Center 2012). In order to go beyond these two cases, our proposal consists of the cross-fertilization of Information Visualization (InfoVis), Library and Information Science (LIS) and Semantic Web disciplines in the definition of alternative representations to images.
 
2 State of the art
 
The theoretical principles that regulate the creation of alternative representation of images could be defined by the combination of models for the analysis of images offered by the discipline of InfoVis and LIS. This analysis can integrate the models of layers of image proposed by LIS (Panofsky 1993) with the rules of composition proposed by InfoVis (Engelhardt 2006). This could be a significant step towards the definition of the image in relation to: the syntactic and semantic visual features; the function in the context in which is presented; the interpretation of its content by the user and the process in which the visual information is created and managed. The visual representation could be considered into the standard process of creation of knowledge and the alternatives viewed as parallel, previous or posterior steps to it (Card et al. 1999).
 
The methods and techniques supplied by LIS and Semantic Web could improve the standardisation and even the automation of the creation and management of alternative representations of images (Kawanaka et al. 2009) (Fredj & Duce 2007). They could facilitate its integration into the publishing workflow improving, at the same time, the retrieval and the adaptation of the images to different contexts of access. A formal description would also benefit the reuse of the information included in the visual representation, for example in the process of cataloguing and retrieval of images.
 
The methodology used in this work is the result of a thorough literature review of the principal theories, models, guidelines and best practices about the description of images in the four disciplines cited, focusing the attention on the need to improve the accessibility, the generation of alternative representations and the retrieval of images.
 
3 R & D work and results / Impact or contributions to the field
 
The current results of the investigation consist of the identification and synthesis of strengths and opportunities of improvements in the accessibility discipline and in the synthesis of the major contributions supplied by the disciplines of InfoVis, LIS and Semantic Web. The strengths and opportunities detected in the accessibility discipline, when analyzed in conjunction with the other disciplines, are presented in Table 1.
 
Strengths Opportunities of improvement
  • Praxis and Guidelines
  • Normative/legal framework
  • Multimodal approach to the description of images
  • Techniques for the inclution of alternative reprentations
  • Specoal attention to the profile of the user, the equipmnet used of the user, the equipment used to access the content and the general context of access
  • Theory to back the process of description
  • Sell lateral benefits, made clear by the other disciplines (retrieval understanding, reuse...).
  • True iclusion of the alternatives into the publishing workflow.

 

[Table 1] Strengths and opportunities detected in the accessibility discipline

 

The expected impact to include the other disciplines methods will be:
 
  • A theory to attain the process of description
  • Standards for the different areas covered in the description
  • Richer techniques to include the description within the image
  • Reuse of the descriptions
  • Increased use of descriptions due to a perceived benefit of retrieval
  • Richer descriptions
  • Automation of part of the process
 
4 Conclusion and planned activities
 
As we have seen InfoVis, LIS and Semantic Web can enhance image description and convert it into a truly implemented activity within the authoring process of a digital document. Starting from this hypothesis we should validate it making specific proposals of integration, analyzing the workflow in the publishing processes, talking to stakeholders, prototyping and evaluating a first model of implementation of the uncovered ideas. As a final result we expect to be able to propose guidelines, models and the implementation of methods and techniques for the description of images and their integration into the publishing workflow, with the aim to standardize and even automate the task of creating and managing alternative representations of images.
 
References
 
Card, S. K., Mackinlay, J. D., Shneiderman, B. Readings in information visualization: using vision to think. Morgan Kaufmann, 1999.
DIAGRAM Center. Digital Image and Graphic Resources for Accessible Materials. 2012. [Accessed October 20, 2011]. Available at: http://diagramcenter.org/.
Engelhardt, Y. Objects and spaces: The visual language of graphics. Diagrammatic Representation and Inference. 2006. p. 104–108.
Fredj, Z. B., Duce, D. A. GraSSML: accessible smart schematic diagrams for all. Universal Access in the Information Society, 6(3). 2007. p. 233–247.
Kawanaka, S. et al. Accessibility commons: a metadata repository for web accessibility. ACM SIGWEB Newsletter. 2009. p. 1:1–1:7.
NCAM. Effective Practices for Description of Science Content within Digital Talking Books. 2009. [Accessed January 22, 2012]. Available at: http://ncam.wgbh.org/experience_learn/educational_media/stemdx/intro.
Panofsky, E. Meaning in the visual art. London: Penguin, 1993.
Petrie, H., Harrison, C., Dev, S. Describing images on the web: a survey of current practice and prospects for the future. Proceedings of Human Computer Interaction International (HCII). 2005.

 

For the full paper and presentation material as well as the record of the presentation please sign in.

top