Semantic Scholar

The non-profit Allen Institute for Artificial Intelligence (AI2) just unveiled an academic database called Semantic Scholar, that they claim can actually understand and interpret the contents of the papers. One of the innovative features of Semantic Scholars is that it allows you to pick out the most important keywords and phrases from the text without relying on an author or publisher to key them in. The system can identify which of a paper's cited references were truly influential, rather than being included incidentally for background or comparison purposes. It also extracts figures and calculations from the papers to present in the search results. Ultimately, the goal of Semantic Scholar is to create a computerized service that can interpret scientific literature and identify useful and relevant hypotheses and experiments.

Similar Databases

Google Scholar is currently the most encompassing, with around 100 million or more documents. However it presents several problems, including the fact that a significant proportion of its documents are considered “unscholarly.” Another is Microsoft Academic Search, which contained over 30 million documents before Microsoft stopped adding to it. Currently, Semantic Scholar can only access 3 million open-access papers in computer science, but AI2 intends to broaden this to other fields within the first year. The AI2 team is particularly interested in expanding into medicine and creating a more efficient and effective structure when it comes to searching through medical research.

Share This Article