Microsoft Academic Graph: When experts are not enough

Abstract

An ongoing project explores the extent to which artificial intelligence (AI), specifically in the areas of natural language processing and semantic reasoning, can be exploited to facilitate the studies of science by deploying software agents equipped with natural language understanding capabilities to read scholarly publications on the web. The knowledge extracted by these AI agents is organized into a heterogeneous graph, called Microsoft Academic Graph (MAG), where the nodes and the edges represent the entities engaging in scholarly communications and the relationships among them, respectively. The frequently updated data set and a few software tools central to the underlying AI components are distributed under an open data license for research and commercial applications. This paper describes the design, schema, and technical and business motivations behind MAG and elaborates how MAG can be used in analytics, search, and recommendation scenarios. How AI plays an important role in avoiding various biases and human induced errors in other data sets and how the technologies can be further improved in the future are also discussed.

Google Scholar, Microsoft Academic, Scopus, Dimensions, Web of Science, and OpenCitations’ COCI: a multidisciplinary comparison of coverage via citations

“…Although Google Scholar and Microsoft Academic are the two most comprehensive bibliographic data sources analysed in this study, their search functionalities have a number of limitations, such as limited support of Boolean and other types of search operators, limited filtering capabilities (Google Scholar), and non-transparent algorithms to process queries and rank the documents in the results page (Microsoft Academic uses artificial intelligence, and Google Scholar uses publicly unknown heuristics to rank documents by relevance) (Beel and Gipp 2009a , b , c ; Martin-Martin et al 2017 ; Orduña-Malea et al 2016 ; Rovira et al 2019 ; Wang et al 2020 ). These characteristics, which prevent users from being able to generate complex search equations that are guaranteed to stay reproducible over time, have led some authors to consider Google Scholar and Microsoft Academic inadequate for query-based search (Gusenbauer and Haddaway 2020 ).…”

“…In 2016, Microsoft launched a new platform called Microsoft Academic, based on Bing’s web crawling infrastructure. Like Google Scholar, Microsoft Academic is a free academic search engine, but unlike Google Scholar, Microsoft Academic facilitates bulk access to its data via an Applications Programming Interface (API) (Wang et al 2020 ).…”

“…Microsoft Academic has been recently reported to cover over 225 million publications (Wang et al 2020 ). Harzing carried out an analysis of her own publication record and the publication records of 145 academics in five broad disciplinary areas (Harzing 2016a , b ; Harzing and Alakangas 2017a , b ).…”

GPT-GNN

“…Open Academic Graph (OAG) [34,38,44] contains more than 178 million nodes and 2.236 billion edges. It is the largest publicly available heterogeneous academic dataset to date.…”

ResearchFlow: Understanding the Knowledge Flow Between Academia and Industry

“…Typically, the two sectors are either analysed separately [15,[17][18][19][20] or together on a small scale [10,11], using a limited sample of papers and patents. Most of these analyses rely on knowledge graphs describing research publications, such as Microsoft Academic Graph [21], Scopus 2 , Semantic Scholar 3 , Aminer [22], Core [23], OpenCitations [24], and others. Other resources, such as Dimensions 4 , the United States Patent and Trademark Office corpus 5 , the PatentScope corpus 6 and the European Patent Office dataset 7 , offer a similar description of patents.…”