IMPORTANT NOTE: This Service is in development. There is a demonstrator of the service available: https://denver.hesge.ch/BiotXplorer/
BiotXplorer is a service designed to search for biotic interactions in the scientific literature. Its primary functions include identifying species that may interact with a given species, determining the type of biotic interaction between two species, and retrieving evidence supporting a biotic interaction between two species. The service works by preprocessing literature to build triplets, which consist of two species and an optional interaction. To refine its system, BiotXplorer has implemented specific rules, such as ensuring the two species in a triplet do not have a parent/child relationship. These triplets are stored in a MongoDB database, accessible via an API and a graphical user interface (GUI).
The service aims to address the challenge of managing the vast volume of data in biodiversity research, which often obstructs valuable insights. Its goal is to assist in detecting biotic interactions, and it processes a large amount of data, including abstracts, full texts, supplementary materials, and PLAZI treatments. The evidence included with each triplet allows domain experts to validate predictions.
BiotXplorer is unique as there are no other automatic builders for biotic interactions currently available. It represents a significant upgrade in biodiversity data usage by enabling connections between databases, especially curated data items. The system can link various types of literature, including abstracts, full-text articles, and supplementary data files, providing a broad view of biotic interactions.
Using BiotXplorer does not require specific competencies, as it offers a user-friendly interface and API access. The service provides query examples to help users get started. However, it's important to note that due to the large volume of data processed, the service might be slow to respond to some queries, and the current development stage includes several false positives that require manual evidence scanning. Efforts are underway to assess and improve the quality of the triplets generated by the system.
Here's a detailed and practical example of how it could be used:
- Preprocessing Literature to Build Triplets: BiotXplorer processes scientific literature to create 'triplets', which consist of two species and an optional interaction between them. This helps in understanding the relationship between various species.
- Querying Biotic Interactions: Suppose a researcher is studying the interactions between specific bird species and their prey. Using BiotXplorer, they can input the name of the bird species and the service will identify potential prey species that may interact with it. The system will determine the nature of the interaction (e.g., predator-prey, symbiotic) and provide relevant literature evidence supporting this interaction.
- Connecting Diverse Data Sources: The service can link different types of literature, including abstracts, full-text articles, and supplementary data files, to provide a comprehensive view of the biotic interactions. This extensive connection of databases and curated data items is a significant advantage for biodiversity research.
- User-Friendly Interface and API Access: BiotXplorer offers a user-friendly GUI and API access, making it accessible to users without specific technical competencies. This allows researchers to easily input their queries and receive valuable insights without needing to navigate through complex databases manually.
- Addressing Biodiversity Research Challenges: By processing large amounts of data and providing validated predictions, BiotXplorer assists in overcoming the challenges posed by the vast volume of data in biodiversity research. This helps in unveiling valuable insights that might otherwise be obscured by the sheer amount of available information.
Last modified: Sunday, 19 November 2023, 9:36 PM