Mohd Hafizul Afifi Abdullah

Purpose: The purpose of this study is to serve as a comprehensive review of the existing annotated corpora. This review study aims to provide information on the existing annotated corpora for event extraction, which are limited but essential for training and improving the existing event extraction algorithms. In addition to the primary goal of this study, it provides guidelines for preparing an annotated corpus and suggests suitable tools for the annotation task. Design/methodology/approach: This study employs an analytical approach to examine available corpus that is suitable for event extraction tasks. It offers an in-depth analysis of existing event extraction corpora and provides systematic guidelines for researchers to develop accurate, high-quality corpora. This ensures the reliability of the created corpus and its suitability for training machine learning algorithms. Findings: Our exploration reveals a scarcity of annotated corpora for event extraction tasks. In particular, the English corpora are mainly focused on the biomedical and general domains. Despite the issue of annotated corpora scarcity, there are several high-quality corpora available and widely used as benchmark datasets. However, access to some of these corpora might be limited owing to closed-access policies or discontinued maintenance after being initially released, rendering them inaccessible owing to broken links. Therefore, this study documents the available corpora for event extraction tasks. Research limitations: Our study focuses only on well-known corpora available in English and Chinese. Nevertheless, this study places a strong emphasis on the English corpora due to its status as a global lingua franca, making it widely understood compared to other languages. Practical implications: We genuinely believe that this study provides valuable knowledge that can serve as a guiding framework for preparing and accurately annotating events from text corpora. It provides comprehensive guidelines for researchers to improve the quality of corpus annotations, especially for event extraction tasks across various domains. Originality/value: This study comprehensively compiled information on the existing annotated corpora for event extraction tasks and provided preparation guidelines.

The novel Deep Neuro-Fuzzy System (DNFS) has attracted significant attention from researchers due to the model’s adaptability and rule-based structure. The model has been successfully implemented in various real-world applications. However, despite its success, the DNFS experience difficulties applying high-dimensional data. The rule base in the DNFS expands exponentially with the number of features affecting the transparency of the model. Additionally, the typical gradient-descent (GD) technique used in the DNFS rule-base optimization frequently encounters the issue of being trapped in local minima. This study aims to introduce a modern optimization approach to address these drawbacks. Therefore, the novel Bitwise Arithmetic Optimization Algorithm (BAOA) has been proposed in this work. The BAOA method has been implemented as a feature selection approach to solve the large rule base problem due to applying high dimensional data. Moreover, the DNFS’ rule base optimization was carried out using the proposed BAOA algorithm to escape the local minima issue owing to the GD algorithm. The simulation results obtained on twelve benchmark datasets illustrated that the BAOA has been able to select the least number of features from high-dimensional data with an average accuracy of 95.53%. In terms of rule base optimization, the novel BAOA achieved better performance with average training and testing accuracies of 96.87% and 96.25%, respectively, on benchmark datasets compared to the Arithmetic Optimization Algorithm (AOA) (average training and testing accuracies of 95.66% and 94.54%) and GD-based optimization (average training and testing accuracies of 94.07% and 93.19%). Additionally, the Wilcoxon test revealed a significant difference between the performances of the proposed BAOA algorithm and comparative methods. The findings indicate that the proposed BAOA approach is highly effective for high-dimensional real-world problems.

Information extraction (IE) is a challenging task, particularly when dealing with highly heterogeneous data. State-of-the-art data mining technologies struggle to process information from textual data. Therefore, various IE techniques have been developed to enable the use of IE for textual data. However, each technique differs from one another because it is designed for different data types and has different target information to be extracted. This study investigated and described the most contemporary methods for extracting information from textual data, emphasizing their benefits and shortcomings. To provide a holistic view of the domain, this comprehensive systematic literature review employed a systematic mapping process to summarize studies published in the last six years (from 2017 to 2022). It covers fundamental concepts, recent approaches, applications, and trends, in addition to challenges and future research prospects in this domain area. Based on an analysis of 161 selected studies, we found that the state-of-the-art models employ deep learning to extract information from textual data. Finally, this study aimed to guide novice and experienced researchers in future research and serve as a foundation for this research area.

In the world we live today, data is the new oil. Data can reveal hidden knowledge that gives us an advantage over our competitors. However, data that are present in an unstructured form such as text documents are difficult to be processed by conventional machine learning algorithms. Therefore, in this study, we attempted to perform information extraction from textual data using current and state-of-the-art models to understand their working mechanisms. To perform this study, we have chosen the GENIA corpus for evaluating the performance of each model. These selected event extraction models are evaluated based on specific measures which are precision, recall, and F-1 measure. The result of our study shows that the DeepEventMine model has scored the highest for trigger detection with a precision of 79.17%, recall at 82.93%, and F-1 measure at 81.01%. Similarly, for event detection, the DeepEventMine model has scored highest among other models with a precision of 65.24%, recall at 55.93%, and F-1 measure at 60.23% based on the selected corpus.