About this Research Topic
A broad spectrum of knowledge about chemistry and biomedicine is captured in textual data, such as scientific publications and patents. Such textual data is rapidly increasing in volume, making it difficult for researchers to identify and track new discoveries and advances. Therefore, developing tools that automatically extract knowledge from these data has attracted extensive attention.
In this research topic, we hope to explore approaches for data mining from unstructured data (e.g., free texts) and semi-structured data (e.g., tables), with emphasis on scientific texts, including bio-chemical and medical data such as genetic or biomolecular information. The nominated research themes include but are not limited to:
• natural language processing methods for scientific text mining and analysis
• novel large language model architectures for scientific text mining and analysis
• evaluation of the efficacy of large language models for (bio)chemical and medical data mining
• construction of knowledge bases or knowledge graphs from scientific texts
• development of novel representation techniques for (bio)chemical entities and concepts
• methods for information extraction such as named entity recognition and identification of relations between entities, in scientific texts
• document-level or multi-document summarization of scientific texts
• multi-modal data mining, such as information alignment between textual data and images
• hybrid knowledge-based/semantic and statistical models for scientific text mining and analysis
• literature-based discovery for scientific hypothesis generation
• systematic review automation methods
Methods that target the identification and extraction of crucial information, such as characteristics of chemicals, polymers, drug names, and molecules are welcome.
Methods that address the linguistic characteristics of specialized biochemical and medical data, such as developing effective representations for large language models, are also welcome.
Research that focuses on resource development such as annotated corpora or domain-specific terminologies, or methods for constituent components of a text mining system, including specialized domain-specific tokenization or chemical structure analysis, are also in scope.
In this research topic, we hope to explore approaches for data mining from unstructured data (e.g., free texts) and semi-structured data (e.g., tables), with emphasis on scientific texts, including bio-chemical and medical data such as genetic or biomolecular information. The nominated research themes include but are not limited to:
• natural language processing methods for scientific text mining and analysis
• novel large language model architectures for scientific text mining and analysis
• evaluation of the efficacy of large language models for (bio)chemical and medical data mining
• construction of knowledge bases or knowledge graphs from scientific texts
• development of novel representation techniques for (bio)chemical entities and concepts
• methods for information extraction such as named entity recognition and identification of relations between entities, in scientific texts
• document-level or multi-document summarization of scientific texts
• multi-modal data mining, such as information alignment between textual data and images
• hybrid knowledge-based/semantic and statistical models for scientific text mining and analysis
• literature-based discovery for scientific hypothesis generation
• systematic review automation methods
Methods that target the identification and extraction of crucial information, such as characteristics of chemicals, polymers, drug names, and molecules are welcome.
Methods that address the linguistic characteristics of specialized biochemical and medical data, such as developing effective representations for large language models, are also welcome.
Research that focuses on resource development such as annotated corpora or domain-specific terminologies, or methods for constituent components of a text mining system, including specialized domain-specific tokenization or chemical structure analysis, are also in scope.
Keywords: Chemical Text, Biochemical Text, Text Mining, Unstructured Natural Language Descriptions, Information Extraction
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
