Constructing a biodiversity terminological inventory

Building a specialized lexicon for breast cancer clinical trial subject eligibility analysis

A natural language processing (NLP) application requires sophisticated lexical resources to support its processing goals. Different solutions, such as dictionary lookup and MetaMap, have been proposed in the healthcare informatics literature to identify disease terms with more than one word (multi-gram disease named entities). Although a lot of work has been done in the identification of protein- and gene-named entities in the biomedical field, not much research has been done on the recognition and resolution of terminologies in the clinical trial subject eligibility analysis. In this study, we develop a specialized lexicon for improving NLP and text mining analysis in the breast cancer domain, and evaluate it by comparing it with the Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT). We use a hybrid methodology, which combines the knowledge of domain experts, terms from multiple online dictionaries, and the mining of text from sample clinical trials. Use of our methodology introduces 4243 unique lexicon items, which increase bigram entity match by 38.6% and trigram entity match by 41%. Our lexicon, which adds a significant number of new terms, is very useful for matching patients to clinical trials automatically based on eligibility matching. Beyond clinical trial matching, the specialized lexicon developed in this study could serve as a foundation for future healthcare text mining applications.

Biodiversity Observations Miner: A web application to unlock primary biodiversity data from published literature

Background

A considerable portion of primary biodiversity data is digitally locked inside published literature which is often stored as pdf files. Large-scale approaches to biodiversity science could benefit from retrieving this information and making it digitally accessible and machine-readable. Nonetheless, the amount and diversity of digitally published literature pose many challenges for knowledge discovery and retrieval. Text mining has been extensively used for data discovery tasks in large quantities of documents. However, text mining approaches for knowledge discovery and retrieval have been limited in biodiversity science compared to other disciplines.

New information

Here, we present a novel, open source text mining tool, the Biodiversity Observations Miner (BOM). This web application, written in R, allows the semi-automated discovery of punctual biodiversity observations (e.g. biotic interactions, functional or behavioural traits and natural history descriptions) associated with the scientific names present inside a corpus of scientific literature. Furthermore, BOM enable users the rapid screening of large quantities of literature based on word co-occurrences that match custom biodiversity dictionaries. This tool aims to increase the digital mobilisation of primary biodiversity data and is freely accessible via GitHub or through a web server.

COPIOUS: A gold standard corpus of named entities towards extracting species occurrence from biodiversity literature

Background Species occurrence records are very important in the biodiversity domain. While several available corpora contain only annotations of species names or habitats and geographical locations, there is no consolidated corpus that covers all types of entities necessary for extracting species occurrence from biodiversity literature. In order to alleviate this issue, we have constructed the COPIOUS corpus-a gold standard corpus that covers a wide range of biodiversity entities. Results Two annotators manually annotated the corpus with five categories of entities, i.e. taxon names, geographical locations, habitats, temporal expressions and person names. The overall inter-annotator agreement on 200 doubly-annotated documents is approximately 81.86% F-score. Amongst the five categories, the agreement on habitat entities was the lowest, indicating that this type of entity is complex. The COPIOUS corpus consists of 668 documents downloaded from the Biodiversity Heritage Library with over 26K sentences and more than 28K entities. Named entity recognisers trained on the corpus could achieve an F-score of 74.58%. Moreover, in recognising taxon names, our model performed better than two available tools in the biodiversity domain, namely the SPECIES tagger and the Global Name Recognition and Discovery. More than 1,600 binary relations of Taxon-Habitat, Taxon-Person, Taxon-Geographical locations and Taxon-Temporal expressions were identified by applying a pattern-based relation extraction system to the gold standard. Based on the extracted relations, we can produce a knowledge repository of species occurrences. Conclusion The paper describes in detail the construction of a gold standard named entity corpus for the biodiversity domain. An investigation of the performance of named entity recognition (NER) tools trained on the gold standard revealed that the corpus is sufficiently reliable and sizeable for both training and evaluation purposes. The corpus can be further used for relation extraction to locate species occurrences in literature-a useful task for monitoring species distribution and preserving the biodiversity.

OpenBiodiv-O: ontology of the OpenBiodiv knowledge management system

BACKGROUND

The biodiversity domain, and in particular biological taxonomy, is moving in the direction of semantization of its research outputs. The present work introduces OpenBiodiv-O, the ontology that serves as the basis of the OpenBiodiv Knowledge Management System. Our intent is to provide an ontology that fills the gaps between ontologies for biodiversity resources, such as DarwinCore-based ontologies, and semantic publishing ontologies, such as the SPAR Ontologies. We bridge this gap by providing an ontology focusing on biological taxonomy.

RESULTS

OpenBiodiv-O introduces classes, properties, and axioms in the domains of scholarly biodiversity publishing and biological taxonomy and aligns them with several important domain ontologies (FaBiO, DoCO, DwC, Darwin-SW, NOMEN, ENVO). By doing so, it bridges the ontological gap across scholarly biodiversity publishing and biological taxonomy and allows for the creation of a Linked Open Dataset (LOD) of biodiversity information (a biodiversity knowledge graph) and enables the creation of the OpenBiodiv Knowledge Management System. A key feature of the ontology is that it is an ontology of the scientific process of biological taxonomy and not of any particular state of knowledge. This feature allows it to express a multiplicity of scientific opinions. The resulting OpenBiodiv knowledge system may gain a high level of trust in the scientific community as it does not force a scientific opinion on its users (e.g. practicing taxonomists, library researchers, etc.), but rather provides the tools for experts to encode different views as science progresses.

CONCLUSIONS

OpenBiodiv-O provides a conceptual model of the structure of a biodiversity publication and the development of related taxonomic concepts. It also serves as the basis for the OpenBiodiv Knowledge Management System.

Temporal degradation of data limits biodiversity research

Spatial and/or temporal biases in biodiversity data can directly influence the utility, comparability, and reliability of ecological and evolutionary studies. While the effects of biased spatial coverage of biodiversity data are relatively well known, temporal variation in data quality (i.e., the congruence between recorded and actual information) has received much less attention. Here, we develop a conceptual framework for understanding the influence of time on biodiversity data quality based on three main processes: (1) the natural dynamics of ecological systems—such as species turnover or local extinction; (2) periodic taxonomic revisions, and; (3) the loss of physical and metadata due to inefficient curation, accidents, or funding shortfalls. Temporal decay in data quality driven by these three processes has fundamental consequences for the usage and comparability of data collected in different time periods. Data decay can be partly ameliorated by adopting standard protocols for generation, storage, and sharing data and metadata. However, some data degradation is unavoidable due to natural variations in ecological systems. Consequently, changes in biodiversity data quality over time need be carefully assessed and, if possible, taken into account when analyzing aging datasets.