We present a manually annotated corpus, Species-Species Interaction, for
extracting meaningful binary relations between species, in biomedical texts, at
sentence level, with a focus on the gut microbiota. The corpus leverages
PubTator to annotate species in full-text articles after evaluating different
Named Entity Recognition species taggers. Our first results are promising for
extracting relations between species using BERT and its biomedical variants.

我们提出了一个手动注释的语料库，物种物种相互作用，用于在句子级别上从生物医学文本中提取物种之间有意义的二元关系，重点关注肠道微生物群落。该语料库利用 PubTator 对全文文章中的物种进行注释，并评估了不同的命名实体识别物种标记器。我们的第一批结果表明，使用 BERT 及其生物医学变体可以很好地提取物种之间的关系。

构建生物医学物种提及关系提取语料库

Building a Corpus for Biomedical Relation Extraction of Species Mentions

Interval-valued information systems are generalized models of single-valued
information systems. By rough set approach, interval-valued information systems
have been extensively studied. Authors could establish many binary relations
from the same interval-valued information system. In this paper, we do some
researches on comparing these binary relations so as to provide numerical
scales for choosing suitable relations in dealing with interval-valued
information systems. Firstly, based on similarity degrees, we compare the most
common three binary relations induced from the same interval-valued information
system. Secondly, we propose the concepts of transitive degree and cluster
degree, and investigate their properties. Finally, we provide some methods to
compare binary relations by means of the transitive degree and the cluster
degree. Furthermore, we use these methods to analyze the most common three
relations induced from Face Recognition Dataset, and obtain that $RF_{B}
^{\lambda}$ is a good choice when we deal with an interval-valued information
system by means of rough set approach.

本研究利用粗糙集方法， 比较区间值信息系统中常见的三种二元关系的相似度，并提出传递度和聚类度的概念，最终通过这些方法对人脸识别数据集中的三种关系进行了分析。