||Architectural Brand valuations through a tag-based learning machine
||Brand is a set of associations related to an object from a particular source (Libai 2008). Such an object can be a product, person or service. Architecture is both service and product. However architectural Branding has never been clearly defined. This probably happens due to the lack of tools to measure the architectural Brand. This paper explores a direction to evaluate the architectural Brand by using computational methods in order to offer a better understanding regarding the awareness, reputation and prominence of the architectural firm. The methodology is based on case studies in which the brands of different types of architectural firms are analyzed, measured and compared to each other through a two-part process: a. the development of one tool to statistically measure the architectural Brand and b. the interpretation of the results of the measurements. a. Learning machine In order to make these brand measurements within a specific group of people or images, this paper develops an interactive tool that uses an image database. The tool constitutes a learning machine; it makes a hypothetic hierarchical categorization of the image database into + and – using an initial sample and it proposes to the user the first image of the list; finally, the user validates the image by confirming or not the machine’s guess. In this image database, each image is described as an array of attributes (tags). Tags can be generated either by the architectural firm itself or can emerge by users. b. Evaluation of results This interactive, user-friendly tool is drawing a user’s preference by proposing similar images from the database based on a learning process from the user input (initial sample and feedback); alternatively, it can be used as a questionnaire for quantitative research. Moreover, this tool categorizes photos of similar content. This research focuses on the following issues as parameters of the effectiveness of the process: o Simplicity of the database organization through computation. o Top-down Vs. Bottom-up tagging of works/ images mechanism. o Future use of the image database. o Transformation of the image database while becoming larger. o “Market” segmentation or not. o Combination of the tool with other Brand measurement tools. o Combination of the tool with other image databases. The outcome of this approach can provide an analysis and metric of the brand strength of different architectural firms. Furthermore, it can help architectural firms to understand better how they are perceived by others in order to improve their brand image and associations.
|Year of publication:
||Branding, learning machine, image database, attribute, tag
Toloudi, Zenovia (2008).
Architectural Brand valuations through a tag-based learning machine. SIGraDi 2008 - [Proceedings of the 12th Iberoamerican Congress of Digital Graphics] La Habana - Cuba 1-5 December 2008,