Spatial Aggregation Facilitates Discovery of Spatial Topics

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 252–262 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 252 Spatial Aggregation Facilitates Discovery of Spatial Topics Aniruddha Maiti Temple University Philadelphia, PA-19122, USA aniruddha.maiti@temple.edu Slobodan Vucetic Temple University Philadelphia, PA-19122, USA vucetic@temple.edu Abstract Spatial aggregation refers to merging of doc- uments created at the same spatial location. We show that by spatial aggregation of a large collection of documents and applying a tra- ditional topic discovery algorithm on the ag- gregated data we can efficiently discover spa- tially distinct topics. By looking at topic dis- covery through matrix factorization lenses we show that spatial aggregation allows low rank approximation of the original document-word matrix, in which spatially distinct topics are preserved and non-spatial topics are aggre- gated into a single topic. Our experiments on synthetic data confirm this observation. Our experiments on 4.7 million tweets collected during the Sandy Hurricane in 2012 show that spatial and temporal aggregation allows rapid discovery of relevant spatial and temporal top- ics during that period. Our work indicates that different forms of document aggregation might be effective in rapid discovery of vari- ous types of distinct topics from large collec- tions of documents. 1 Introduction Social microblogging sites such as Twitter gener- ate large volumes of short documents through the activity of hundreds of millions of users around the world. This provides an unprecedented ac- cess to the pulse of the global society. Due to the sheer volume and diversity of the generated con- tent, topic discovery has been an invaluable tool in an effort to make sense of this data. Regardless of a precise definition of a topic and a particular topic model, topics discovery is used to describe per- tinent themes in a document corpus and serve to identify events, trends, and interests at the global, local, or a social group level. Among the most popular topic modeling tech- niques are Latent Dirichlet Allocation (LDA), La- tent Semantic Analysis (LSA), and Non-negative Matrix factorization (NMF). When applying those techniques for topic discovery from microblogs, there are three main challenges: (1) how to im- prove computational speed, (2) how to extract use- ful topics, and (3) how to deal with short texts. Many papers were published that address one or more of these challenges and most of them pro- pose to modify the original topic models. In this paper, we are focusing on aggregation (also referred to as pooling) (Alvarez-Melis and Saveski, 2016) (Hong and Davison, 2010) (Weng et al., 2010) (Steinskog et al., 2017), a particular document preprocessing technique that has been empirically shown to be useful for topic discov- ery from microblogs. The main idea of aggrega- tion is to combine multiple documents into a sin- gle document according to some external criterion and to apply a topic discovery algorithm on the ag- gregated documents. The earliest mentions of ag- gregation (Mehrotra et al., 2013) (Hong and Davi- son, 2010)(Weng et al., …

Content truncated to 500 words