Background Chemical and biomedical named entity recognition (NER) is an essential preprocessing task in natural language processing . The identification and extraction of named entities from scientific articles is also attracting increasing interest in many scientific disciplines. Locating chemical named entities in the literature is an essential step in chemical text mining pipelines for identifying chemical mentions, their properties, and relations as discussed in the literature. In this work, we describe an approach to the BioCreative V.5 challenge regarding the recognition and classification of chemical named entities. For this purpose, we transform the task of NER into a sequence labeling problem. We present a series of sequence labeling systems that we used, adapted and optimized in our experiments for solving this task. To this end, we experiment with hyperparameter optimization. Finally, we present LSTMVoter, a two-stage application of recurrent neural network s that integrates the optimized sequence labelers from our study into a single ensemble classifier. Results We introduce LSTMVoter, a bidirectional long short-term memory (LSTM) tagger that utilizes a conditional random field layer in conjunction with attention-based feature modeling. Our approach explores information about features that is modeled by means of an attention mechanism. LSTMVoter outperforms each extractor integrated by it in a series of experiments. On the BioCreative IV chemical compound and drug name recognition (CHEMDNER) corpus, LSTMVoter achieves an F1-score of 90.04%; on the BioCreative V.5 chemical entity mention in patents corpus, it achieves an F1-score of 89.01%. Availability and implementation Data and code are available at https://github.com/texttechnologylab/LSTMVoter .

Are nearby places (e.g. cities) described by related words? In this article we transfer this research question in the eld of lexical encoding of geographic information onto the level of intertextuality. To this end, we explore Volunteered Geographic Information (VGI) to model texts addressing places at the level of cities or regions with the help of so-called topic networks. is is done to examine how language encodes and networks geographic information on the aboutness level of texts. Our hypothesis is that the networked thematizations of places are similar -regardless of their distances and the underlying communities of authors. To investigate this we introduce Multiplex Topic Networks (MTN), which we automatically derive from Linguistic Multilayer Networks (LMN) as a novel model, especially of thematic networking in text corpora. Our study shows a Zip an organization of the thematic universe in which geographical places (especially cities) are located in online communication. We interpret this nding in the context of cognitive maps, a notion which we extend by so-called thematic maps. According to our interpretation of this nding, the organization of thematic maps as part of cognitive maps results from a tendency of authors to generate shareable content that ensures the continued existence of the underlying media. We test our hypothesis by example of special wikis and extracts of Wikipedia. In this way we come to the conclusion: Places, whether close to each other or not, are located in neighboring places that span similar subnetworks in the topic universe.

The recognition of pharmacological substances, compounds and proteins is an essential preliminary work for the recognition of relations between chemicals and other biomedically relevant units. In this paper, we describe an approach to Task 1 of the PharmaCoNER Challenge, which involves the recognition of mentions of chemicals and drugs in Spanish medical texts. We train a state-of-the-art BiLSTM-CRF sequence tagger with stacked Pooled Contextualized Embeddings, word and sub-word embeddings using the open-source framework FLAIR. We present a new corpus composed of articles and papers from Spanish health science journals, termed the Spanish Health Corpus, and use it to train domainspecific embeddings which we incorporate in our model training. We achieve a result of 89.76% F1-score using pre-trained embeddings and are able to improve these results to 90.52% F1-score using specialized embeddings.

The storage of data in public repositories such as the Global Biodiversity Information Facility (GBIF) or the National Center for Biotechnology Information (NCBI) is nowadays stipulated in the policies of many publishers in order to facilitate data replication or proliferation. Species occurrence records contained in legacy printed literature are no exception to this. The extent of their digital and machine-readable availability, however, is still far from matching the existing data volume (Thessen and Parr 2014). But precisely these data are becoming more and more relevant to the investigation of ongoing loss of biodiversity. In order to extract species occurrence records at a larger scale from available publications, one has to apply specialised text mining tools. However, such tools are in short supply especially for scientific literature in the German language. The Specialised Information Service Biodiversity Research*1 BIOfid (Koch et al. 2017) aims at reducing this desideratum, inter alia, by preparing a searchable text corpus semantically enriched by a new kind of multi-label annotation. For this purpose, we feed manual annotations into automatic, machine-learning annotators. This mixture of automatic and manual methods is needed, because BIOfid approaches a new application area with respect to language (mainly German of the 19th century), text type (biological reports), and linguistic focus (technical and everyday language). We will present current results of the performance of BIOfid’s semantic search engine and the application of independent natural language processing (NLP) tools. Most of these are freely available online, such as TextImager (Hemati et al. 2016). We will show how TextImager is tied into the BIOfid pipeline and how it is made scalable (e.g. extendible by further modules) and usable on different systems (docker containers). Further, we will provide a short introduction to generating machine-learning training data using TextAnnotator (Abrami et al. 2019) for multi-label annotation. Annotation reproducibility can be assessed by the implementation of inter-annotator agreement methods (Abrami et al. 2020). Beyond taxon recognition and entity linking, we place particular emphasis on location and time information. For this purpose, our annotation tag-set combines general categories and biology-specific categories (including taxonomic names) with location and time ontologies. The application of the annotation categories is regimented by annotation guidelines (Lücking et al. 2020). Within the next years, our work deliverable will be a semantically accessible and data-extractable text corpus of around two million pages. In this way, BIOfid is creating a new valuable resource that expands our knowledge of biodiversity and its determinants.

We test the hypothesis that the extent to which one obtains information on a given topic through Wikipedia depends on the language in which it is consulted. Controlling the size factor, we investigate this hypothesis for a number of 25 subject areas. Since Wikipedia is a central part of the web-based information landscape, this indicates a language-related, linguistic bias. The article therefore deals with the question of whether Wikipedia exhibits this kind of linguistic relativity or not. From the perspective of educational science, the article develops a computational model of the information landscape from which multiple texts are drawn as typical input of web-based reading. For this purpose, it develops a hybrid model of intra-and intertextual similarity of different parts of the information landscape and tests this model on the example of 35 languages and corresponding Wikipedias. In the way it measures the similarities of hypertexts, the article goes beyond existing approaches by examining their structural and semantic aspects intra-and intertextually. In this way it builds a bridge between reading research, educational science, Wikipedia research and computational linguistics.