The big picture of cities: Analysing Flickr photos of 222 cities worldwide

The road to recovery: Sensing public opinion towards reopening measures with social media data in post-lockdown cities

“…Topic modelling is one of the most powerful text mining tools for exploring semantic structure from a collection of texts. Latent Dirichlet Allocation (LDA), as a generative probabilistic model, is commonly employed to extract topics from a collection of documents ( Capela & Ramirez-Marquez, 2019 ; Taecharungroj & Mathayomchan, 2020 ). In the LDA topic model, the collection of documents is referred as the corpus ; items within the corpus are referred to as the document , with specific words in documents called terms .…”

Traveller‐generated destination image: Analysing Flickr photos of 193 countries worldwide

“…Then, LDA was conducted to identify the underlying topics (cognitive attributes) within a corpus of photos. The combination of label detection and topic-modelling techniques was first used by Taecharungroj and Mathayomchan (2020) to study cities. However, it has not been used at the country level to identify destination images.…”

“…To ensure an equitable representation of countries, the photos of each country were screened to reduce the number to 1,500 per country, which is higher than the numbers used in the previous studies: 900 (Hunter, 2016) and 1,000 (Taecharungroj & Mathayomchan, 2020). To reduce the number of photos, those with the fewest detected labels were removed because they would contribute less to the subsequent LDA, which models CIAs based on the co-occurrence of labels.…”

Using Photographs in Mixed Methods Research: An Illustration Integrating Quantitative and Qualitative Analyses from Consumer Research

“…Quantitative methods and photographs have been used, for example, in e-commerce and advertising research (e.g., Taecharungroj & Mathayomchan, 2020), usability research, user-friendliness and consumer experience research (e.g., Vilnai-Yavetz & Tifferet, 2015), social media analysis (Tsai et al, 2016), and visual marketing (Kleih & Sparke, 2021). Photographs are sourced from consumers (e.g., Facebook and Instagram postings, and Flickr), the Internet, or existing advertising, and are analyzed using different software applications and deep-learning algorithms such as Google Cloud Vision API (e.g., Taecharungroj & Mathayomchan, 2020). Again, researchers consider the content of the photographs used to be less important than their effects on the consumers.…”