Toward Universal Spatialization Through Wikipedia-Based Semantic Enhancement

Abstract

This article introduces Cartograph, a visualization system that harnesses the vast world knowledge encoded within Wikipedia to create thematic maps of almost any data. Cartograph extends previous systems that visualize non-spatial data using geographic approaches. Although these systems required data with an existing semantic structure, Cartograph unlocks spatial visualization for a much larger variety of datasets by enhancing input datasets with semantic information extracted from Wikipedia. Cartograph's map embeddings use neural networks trained on Wikipedia article content and user navigation behavior. Using these embeddings, the system can reveal connections between points that are unrelated in the original datasets but are related in meaning and therefore embedded close together on the map. We describe the design of the system and key challenges we encountered. We present findings from two user studies exploring design choices and use of the system. CCS Concepts: • Human-centered computing → Visualization systems and tools; Wikis;

Wikipedia Reader Navigation: When Synthetic Data Is Enough

“…We showed that synthetic navigation sequences from the public clickstream dataset are effective for many practical applications for all 8 considered languages. We compared the performance of real and synthetic navigation sequences in four different downstream tasks involving the use of navigation sequences ( link prediction [46], next-article prediction [12], generating representations [60], and topic classification [35]), revealing that using clickstream data often yields performance that are within 10% (or less) in comparison to using real navigation sequences (Table S2). Specifically, article embeddings generated from synthetic navigation sequences are of comparable quality to those generated from real navigation sequences.…”

“…Paranjape et al [46] showed that navigation traces provide a strong signal to predict new useful links among Wikipedia articles. Furthermore, reading sessions can be used to construct article embeddings-where similarity captures articles that are read in close succession [75]-that have been used for building alternative visual representations of the semantic space of Wikipedia [60]. Applications of clickstream data.…”

The State of the Art in Map‐Like Visualization

“…Sen et al . [SSL*17, SSL*19] add virtual “water points” to the visualization for generating “water” regions not encircled by the Voronoi mesh for low‐density parts on the spatialized data. In R‐Map [CLCY20] and E‐Map [CCL*17, CCA*18], an initial force‐directed layout is used to compute a Voronoi mesh on graph data.…”

“…Cartograph [SSL*19] for example uses the pin symbol to indicate the location of a data item, similar to digital maps (see Figure 5a). While the symbol also resembles the signal poles used in American football, their wide‐spread use to indicate locations in digital maps makes them recognizable as such for point‐based imitation.…”

“…Thereby, we give an overview of the disparate perspectives on map‐like visualization and exemplify the ubiquity of map‐like approaches in visualization. We found applications of map‐like visualizations for example in geo‐visualization (e.g., cartograms [HKPS05]), tree visualization (e.g., spatially‐ordered treemaps [WD08]), graph visualization (e.g., [NPL*15]), visualization of document collections (e.g., strategic knowledge maps [PDJCRBBM12]), software visualization (e.g., software cartography [KELN10]), bibliometric analysis [Hoo07], and visualization of multi‐variate data (e.g., using t‐sne [SSL*19]).…”