We demonstrate an end-to-end question answering system that integrates BERT with the open-source Anserini information retrieval toolkit. In contrast to most question answering and reading comprehension models today, which operate over small amounts of input text, our system integrates best practices from IR with a BERT-based reader to identify answers from a large corpus of Wikipedia articles in an end-to-end fashion. We report large improvements over previous results on a standard benchmark test collection, showing that fine-tuning pretrained BERT with SQuAD is sufficient to achieve high accuracy in identifying answer spans.
One technique to improve the retrieval effectiveness of a search engine is to expand documents with terms that are related or representative of the documents' content. From the perspective of a question answering system, this might comprise questions the document can potentially answer. Following this observation, we propose a simple method that predicts which queries will be issued for a given document and then expands it with those predictions with a vanilla sequence-to-sequence model, trained using datasets consisting of pairs of query and relevant documents. By combining our method with a highly-effective re-ranking component, we achieve the state of the art in two retrieval tasks. In a latencycritical regime, retrieval results alone (without re-ranking) approach the effectiveness of more computationally expensive neural re-rankers but are much faster.Code to reproduce experiments and trained models can be found at https://github. com/nyu-dl/dl4ir-doc2query.
This paper applies BERT to ad hoc document retrieval on news articles, which requires addressing two challenges: relevance judgments in existing test collections are typically provided only at the document level, and documents often exceed the length that BERT was designed to handle. Our solution is to aggregate sentence-level evidence to rank documents. Furthermore, we are able to leverage passage-level relevance judgments fortuitously available in other domains to fine-tune BERT models that are able to capture cross-domain notions of relevance, and can be directly used for ranking news articles. Our simple neural ranking models achieve state-of-the-art effectiveness on three standard test collections.
In medical diagnoses and treatments, e.g., endoscopy, dosage transition monitoring, it is often desirable to wirelessly track an object that moves through the human GI tract. In this paper, we propose a magnetic localization and orientation system for such applications. This system uses a small magnet enclosed in the object to serve as excitation source, so it does not require the connection wire and power supply for the excitation signal. When the magnet moves, it establishes a static magnetic field around, whose intensity is related to the magnet's position and orientation. With the magnetic sensors, the magnetic intensities in some predetermined spatial positions can be detected, and the magnet's position and orientation parameters can be computed based on an appropriate algorithm. Here, we propose a real-time tracking system developed by a cubic magnetic sensor array made of Honeywell 3-axis magnetic sensors, HMC1043. Using some efficient software modules and calibration methods, the system can achieve satisfactory tracking accuracy if the cubic sensor array has enough number of 3-axis magnetic sensors. The experimental results show that the average localization error is 1.8 mm.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.