Search for a command to run...
This dataset investigates the values provided to users by "Place Reconstructions"—projects that recreate real-world locations in either physical forms (e.g., dioramas, relocated buildings) or virtual environments (e.g., Metaverse, video games, VR). The collection includes 79 reconstruction cases, URLs of 1,003 social media reactions, and 186 values extracted through the KA method. Note on Language: To ensure international accessibility, this dataset provides both the original Japanese data and its English translations. Dataset Contents The dataset consists of the following bilingual files: Values (KA Cards) Created from Interview Data (1. KA Cards from Interview Data .csv and 1. KA Cards from Interview Data (Japanese).csv) It includes 42 values It is based on the value map of Aoyagi et al. 2025 Mid-level Values Created from Interview Data (2. Mid-level Values Created from Interview Data.csv and 2. Mid-level Values Created from Interview Data (Japanese).csv) It includes 62 mid-level values. Place Reconstruction Cases (3. Place Recontruction Cases.csv and 3. Place Recontruction Cases (Japanese).csv) A comprehensive list of 79 analyzed projects, covering various formats such as the Haneda Airport diorama, Virtual Shibuya, Minecraft NYC, and the Sistine Chapel reconstruction at the Otsuka Museum of Arts Social Media Post URL(4. Social Media Post URLs.csv) URLs of 1,003 user posts collected from platforms like X (formerly Twitter), Instagram, YouTube, Steam and Reddit. The posts described user experiences, emotional responses, and episodes related to place reconstructions. Due to copyright considerations, this data does not include texts of the posts. Coding Prompts (5. Coding Prompts.txt and 5. Coding Prompts (Japanese).txt) It orders LLM to coding raw post data to KA Cards (user experience values). Values (KA Cards) Created From Social Media Data (6. KA Cards Created From Social Media Data.csv and 6. KA Cards Created From Social Media Data (Japanese).csv) It includes 186 values. Grouping Prompts (6. Grouping Propmpts.txt and 6. Grouping Propmpts (Japanese).txt It orders LLM to grouping KA Cards (experiencevalues) to Mid-level values. It includes four prompts for step execution.The 1st prompt extracts elements and connections from KA cards, then the 2nd prompt group them into mid-level values. The 3rd prompt makes some fix to the result. The 4th prompot prepared for creation of a value map. Mid-Level Values (8. Mid-Level Values .csv and 8. Mid-Level Values (Japanese).csv) It includes 32 mid-level values. It was organized by Relation to Place and Relation to Place Reconstruction.