We are focused on the branches of AI that help make content more easily and quickly discoverable (and also easier to monetize). Critical to that capability is understanding what users are looking for even before they begin a search.
For example, search augmentation tactics can personalize discovery by gauging a user’s interests and returning search results ranked according to relevance. Our work with recommender systems is refining how Literatum targets content to individual users by synthesizing their online behavior and interests.
Predicting the future
AI-based knowledge representation will enable readers to ask questions and get relevant, meaningful answers rather than searching by keyword. Publication knowledge graphs, which identify and establish connections between many types of publication attributes, will improve Literatum’s content recommendations, predict trending research topics, and even which authors and institutions are becoming more influential.
In fact, Atypon’s history of investments in machine learning, natural language processing, and entity recognition have already contributed to award-winning semantic technologies. Our auto-tagger enables Literatum to dynamically populate topic-specific web pages with relevant content; our implementation of Automatic Topic Modeling analyzes a set of documents, determines the topic of each, and assesses their relatedness.