Home Document Upload DTM Viz Heatmaps
Regen-X

Welcome to our recommender system! Copy and paste the contents of the research paper with topics you are interested in researching. Our dynamic topic modeling algorithm will recommend some papers in our corpus that should relate to the research that you are investigating. The advantage of this recommendation is that it is much more powerful than a traditional Google or Pubmed search query. Our algorithm bridges the gap between changes in diction, syntax, and sentence structure over time. In other words, the language in this paper could be different but it ultimately discusses the same topic as the text you entered. The advantage here is that we could recommend a paper written 100 years earlier with less common words that matches the context of your paper. Additionally, we don't keyword match which is what popular search engines like Google do. We match based on the hellinger distance between the topic distribution of the text you entered and each paper in our corpus. The model you are querying now split our corpus into 10 topics and binned the time years into 16 buckets.