A Movement Data Analysis and Synthesis Tool for Museum Visitors’ Behaviors

Abstract

Abstract. Achievement of museum guide systems, in physical and virtual worlds, providing the personalization and context awareness features requires the prior analysis and identification of visitors' behaviors. This paper analyzes and synthesizes visitors' behaviors in museums and art galleries by using our defined parameters. A visit time and a observation distance can be calculated by using the proposed functions. The proposed synthesis algorithm is developed and used in classification. Classifying visitor styles is simply implemented by using the average and variance of their stopover time at and distance to all exhibits as shown in this paper.

Geo-sociograms

“…Chittaro et al [6] developed a visualization tool and verified these styles even in virtual environments. Sookhanaphibarn et al [10] used this approach to determine parameters that influence the style of a visit, in order to predict user behavior. They also pointed out that the social context of people visiting in groups influences their visit.…”

An associative engines based approach supporting collaborative analytics in the Internet of cultural things

“…In the technical literature, these visiting styles have been simulated by using 100 mathematical functions [2], or detected using clustering approaches (Artificial Neural Networks and K-means algorithm) with the objective of providing a personalized service at the earliest stage possible of the visit [3]. Further studies investigated the effects of mobile multimedia location-aware guides, revealing that visitors using a mobile guide visited the museum longer while reducing 105 social interactions with their fellow group members [4,41].…”

Personal storytelling: Using Natural Language Generation for children with complex communication needs, in the wild…

“…This requires domain knowledge and hence knowledge acquisition. For example, a logging system in a museum can utilise information about the exhibits and model the behavior of typical users (see e.g., [37,38]) to generate relevant stories. Also, it is useful to aggregate low-level sensor information into higher-level concepts which are more meaningful to users; for example Nyx…”