Drug Knowledge Extraction Framework with Entity Pair Calibration for Chinese Drug Instructions

Existing pharmaceutical information extraction research often focus on standalone entity or relationship identification tasks over drug instructions. There is a lack of a holistic solution for drug knowledge extraction. Moreover, current methods perform poorly in extracting fine-grained interaction relations from drug instructions. To solve these problems, this paper proposes an information extraction framework for drug instructions. The framework proposes deep learning models with fine-tuned pre-training models for entity recognition and relation extraction, in addition, it incorporates an novel entity pair calibration process to promote the performance for fine-grained relation extraction. The framework experiments on more than 60k Chinese drug description sentences from 4000 drug instructions. Empirical results show that the framework can successfully identify drug related entities (F1 >= 0.95) and their relations (F1 >= 0.83) from the realistic dataset, and the entity pair calibration plays an important role (~5% F1 score improvement) in extracting fine-grained relations.

Analisis Kohesi, Koherensi, dan Skematik Teks Surat Pembaca Bali Post Terkait Covid-19 Periode Maret-Agustus 2020

This study discusses cohesion, coherence, and schematic in the text of the Bali Post reader's letters. Reader's letters contain short letters written by the public with topics that are in accordance with the current situation and related to the public interest. The choice of topics related to Covid-19 is due to the fact that currently many people are writing their complaints due to the Covid-19 pandemic or expressions of praise related to handling Covid-19 which are conveyed through letters to readers of Bali Post. The reader's letter written by the community is interesting to analyze in terms of cohesion, coherence and schematic. The purpose of this study was to determine the tools of cohesion, coherence and schematic of the text of the Bali Post reader's letter. The methods and techniques of data collection used in this study are the method of listening to the technique of note taking. At the data analysis stage, the distribution method and the matching method were used with the deletion and substitution techniques. The theoretical basis used in this research is discourse theory in which there is cohesion proposed by Halliday and R. Hasan (1976:4), coherence proposed by Kridalaksana (in Mulyana, 2005:32) and schematic/superstructure analysis. proposed by Van Dijk (in Eriyanto, 2012:229). Based on the analysis, the following results were found. First, it was found the use of cohesion tools, both grammatical cohesion consisting of reference, substitution, deletion, and concatenation as well as lexical cohesion consisting of repetition, word equivalents, opposites, synonyms, and equivalence. Second, the elements of coherence found include cause-and-effect relationships, means-result relationships, cause-reason relationships, means-end relationships, conclusions-setting relationships, slack-result relationships, conditional-result relationships, comparative relationships, and relationships paraphrastic, amplification relationship, time/temporal additive relationship, nontime/temporal additive relationship, identification relationship, generic-specific relationship, and like relationship. Third, a schematic of the text was found consisting of a title structure, opening structure, content structure, and closing structure.

Sustainable Development Goal Relational Modelling and Prediction

A methodology for UN Sustainable Development Goal (SDG) attainment prediction is presented, the Sustainable Development Goals Correlation Attainment Predictions Extended framework SDG-CAP-EXT. Unlike previous SDG attainment methodologies, SDG-CAP-EXT takes into account the potential for a causal relationship between SDG indicators both with respect to the geographic entity under consideration (intra-entity) and neighbouring geographic entities to the current entity (inter-entity). The challenge is in the discovery of such causal relationships. A ensemble approach is presented that combines the results of a number of alternative causality relationship identification mechanisms. The identified relationships are used to build multi-variate time series prediction models that feed into a bottom-up SDG prediction taxonomy, which is used to make SDG attainment predictions and rank countries using a proposed Attainment Likelihood Index that reflects the likelihood of goal attainment. The framework is fully described and evaluated. The evaluation demonstrates that the SDG-CAP-EXT framework can produce better predictions than alternative models that do not consider the potential for intra- and inter-causal relationships.

Comparative analyses and phylogenetic relationships of 15 Trapa (Trapaceae) species based on Complete Chloroplast Genomes

Trapa L. is floating-leaved aquatic plants with important economic and ecological values. However, species identification and phylogenetic relationship are still unresolved for Trapa. In this study, complete chloroplast genomes of 13 Trapa species/taxa were sequenced and annotated. Combined with released sequences of the other two species, comparative analysis of cp genomes was first performed on the 15 Trapa species/taxa. The 15 cp genomes exhibited quadripartite structures with medium size of 155, 453-155, 559 bp. IR/SC junctions were conservative with no obvious change found. Long repetitive repeats and SSRs were mostly detected in the intergenic and LSC regions, providing useful plastid markers for species and relationship identification. Three phylogenetic analyses (MP, ML and BI) consistently showed two clusters within Trapa, including large- and small-seed species/taxa, respectively. This study provided the baseline information for phylogeography of Trapa, which would facilitate the management and utilization of genetic resources of the genus.

Automatic Detection of Discrimination Actions from Social Images

In this paper, we developed a practical approach for automatic detection of discrimination actions from social images. Firstly, an image set is established, in which various discrimination actions and relations are manually labeled. To the best of our knowledge, this is the first work to create a dataset for discrimination action recognition and relationship identification. Secondly, a practical approach is developed to achieve automatic detection and identification of discrimination actions and relationships from social images. Thirdly, the task of relationship identification is seamlessly integrated with the task of discrimination action recognition into one single network called the Co-operative Visual Translation Embedding++ network (CVTransE++). We also compared our proposed method with numerous state-of-the-art methods, and our experimental results demonstrated that our proposed methods can significantly outperform state-of-the-art approaches.