id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_nptfrj23brdvvauxagfvg2dske	Alison Smith	Evaluating Visual Representations for Topic Understanding and Their Effects on Manually Generated Topic Labels	2017	16	.pdf	application/pdf	8340	756	59	Although all visualizations produce similar quality labels, simple visualizations such as word lists allow participants to quickly understand topics, while These sets of words or "topics" evince internal coherence and can help guide users to relevant To better understand these problems, we use labeling to evaluate topic model visualizations. topic, while a second set of users assessed the quality of those labels alongside automatically generated of topics, while more complex visualizations (network graph) take longer but reveal relationships between words. Automatic labels are generated from representative Wikipedia article titles using a technique similar to Lau et al. Additionally, the word list, word cloud, and network graph visualizations all lead to labels with similar "best" and "worst" votes for both the top and that the network graph helps users to better understand the topic words as a group and therefore label them using a hypernym.	./cache/work_nptfrj23brdvvauxagfvg2dske.pdf	./txt/work_nptfrj23brdvvauxagfvg2dske.txt