Predicting environmental features by learning spatiotemporal embeddings from social media

Abstract

Spatiotemporal modelling is an important task for ecology. Social media tags have been found to have great potential to assist in predicting aspects of the natural environment, particularly through the use of machine learning methods. Here we propose a novel spatiotemporal embeddings model, called SPATE, which is able to integrate textual information from the photo-sharing platform Flickr and structured scientific information from more traditional environmental data sources. The proposed model can be used for modelling and predicting a wide variety of ecological features such as species distribution, as well as related phenomena such as climate features. We first propose a new method based on spatiotemporal kernel density estimation to handle the sparsity of Flickr tag distributions over space and time. Then, we efficiently integrate the spatially and temporally smoothed Flickr tags with the structured scientific data into low-dimensional vector space representations. We experimentally show that our model is able to substantially outperform baselines that rely only on Flickr or only on traditional sources.

Implementation of Flora and Fauna-Based Smart Word Learning Media to Empower Elementary School Students' Motivation and Science Literacy

“…Jeawak et al, (2020) Sebagai guru profesional harus berusaha mendayagunakan potensi kelas, memfokuskan perhatian kepada peserta didik, memahami mereka secara individu dan memberi pelayanan-pelayanan tertentu yang merupakan wujud dukungan untuk meningkatkan pemahaman mereka. Jeawak et al, (2020) Disamping itu, guru memberikan tugas dan kegiatan peserta didik berupa lembar kerja peserta didi (LKPD) tujuannya agar peserta didik lebih dominan aktif dalam kegiatan pembelajaran, bukan guru yang mendominasi dalam pembelajaran Gosal and Ziv (2020). Upaya yang dilakukan ini merupakan usaha dalam menciptakan kondisi belajar yang kondusif, aktif, kreatif, inovatif, optimal, dan menyenangkan dalam proses pembelajaran.…”

Spatiotemporal data mining: a survey on challenges and open problems

“…The sparsity issue directly affects spatiotemporal social media analysis tasks such as density estimation (Jeawak et al. 2020 ; Sakaki et al. 2010 ), event location extraction Chung et al.…”

Spatiotemporal Data Mining: A Survey on Challenges and Open Problems

“…This topic modelling is affected by challenges related to the heterogeneity of geographical context [114], such as the locations sparsity caused by a tiny amount of posts that are tagged with geographical locations. The sparsity issue directly affects spatiotemporal social media analysis tasks such as density estimation [111,230], event location extraction [40] and collaborative filtering [341]. Such tasks are affected by different social factors such as social trust.…”