Related papers: Improving astroBERT using Semantic Textual Similar…
With the mainstream adoption of references to datasets in astronomical manuscripts, researchers today are able to provide direct links from their papers to the original data that were used in their study. Following a process similar to the…
We discuss current efforts behind the curation of observing proposals, archive bibliographies, and data links in the NASA Astrophysics Data System (ADS). The primary data in the ADS is the bibliographic content from scholarly articles in…
In this review, we describe the application of one of the most popular deep learning-based language models - BERT. The paper describes the mechanism of operation of this model, the main areas of its application to the tasks of text…
Over the next decade we will witness the development of a new infrastructure in support of data-intensive scientific research, which includes Astronomy. This new networked environment will offer both challenges and opportunities to our…
We introduce AnnualBERT, a series of language models designed specifically to capture the temporal evolution of scientific text. Deviating from the prevailing paradigms of subword tokenizations and "one model to rule them all", AnnualBERT…
The SAO/NASA Astrophysics Data System (ADS) grew up with and has been riding the waves of the Information Age, closely monitoring and anticipating the needs of its end-users. By now, all professional astronomers are using the ADS on a daily…
The powerful discovery capabilities available in the ADS bibliographic services are possible thanks to the design of a flexible search and retrieval system based on a relational database model. Bibliographic records are stored as a corpus…
Several different controlled vocabularies have been developed and used by the astronomical community, each designed to serve a specific need and a specific group. The Unified Astronomy Thesaurus (UAT) attempts to provide a highly structured…
Scientific literature in astronomy is rapidly expanding, making it increasingly important to automate the extraction of key entities and contextual information from research papers. In this paper, we present an encoder-based system for…
Machine based text comprehension has always been a significant research field in natural language processing. Once a full understanding of the text context and semantics is achieved, a deep learning model can be trained to solve a large…
Being able to extract from scientific papers their main points, key insights, and other important information, referred to here as aspects, might facilitate the process of conducting a scientific literature review. Therefore, the aim of our…
The Smithsonian/NASA ADS Abstract Service contains a wealth of data for astronomers and librarians alike, yet the vast majority of usage consists of rudimentary searches. Hints on how to obtain more focused search results by using more of…
Acronym identification focuses on finding the acronyms and the phrases that have been abbreviated, which is crucial for scientific document understanding tasks. However, the limited size of manually annotated datasets hinders further…
Building multi-modal language models has been a trend in the recent years, where additional modalities such as image, video, speech, etc. are jointly learned along with natural languages (i.e., textual information). Despite the success of…
Over the past decade, astronomers have been using an increasingly larger number of web-based applications and archives to conduct their research. However, despite the early success in creating links across projects and data centers, the…
Large pre-trained language models such as BERT have shown their effectiveness in various natural language processing tasks. However, the huge parameter size makes them difficult to be deployed in real-time applications that require quick…
NASA regards data handling and archiving as an integral part of space missions, and has a strong track record of serving astrophysics data to the public, beginning with the the IRAS satellite in 1983. Archives enable a major science return…
Machine learning algorithms are highly useful for the classification of time series data in astronomy in this era of peta-scale public survey data releases. These methods can facilitate the discovery of new unknown events in most…
This paper presents a novel methodology for improving the performance of machine learning based space traffic management tasks through the use of a pre-trained orbit model. Taking inspiration from BERT-like self-supervised language models…
The academic literature of social sciences records human civilization and studies human social problems. With its large-scale growth, the ways to quickly find existing research on relevant issues have become an urgent demand for…