Header Image

Special Sessions

Get the most out of Biodiversity Knowledge Graphs

Knowledge Graphs (KGs) are used to manage and link information in the age of Big Data. KGs are graph-structured knowledge bases that store factual information in the form of relationships between entities. These graphs consist of nodes and edges connecting them. Nodes represent entities, i.e., concepts and real-world things. Edges represent relationships among entities. Both nodes and edges have unambiguous labels. For example, in the Biodiversity domain, entities could be abstract concepts like a kingdom, a specie, or a trait, or a concrete specimen of a species.Example relationships could be: co-occurs, possesses-trait, collected-at and, etc.

Because of the underlying formal model, there are other important ways KGs can be used:

  • with the help of unique identifiers, KGs can be interlinked with general-purpose KGs like Wikidata or other domain-specific KGs. They are thus an excellent tool for data integration.
  • they can be queried using SPARQL, a powerful query language.
  • new knowledge can be inferred using, e.g., machine learning or data mining. For example, one could infer, that two nodes represent the same real-world entity based on the similarity of the subgraphs surrounding them.
  • they can create the Linked Open Data (LOD) dataset by interlinking cross-domain and domain-specific KGs.

The potential of KGs has been recognized for Biodiversity science, in particular, it can:

  • facilitate a greater alignment between data and expertise in the Biodiversity domain,
  • provide a discovery tool for Biodiversity data,
  • establish a robust infrastructure for managing Biodiversity knowledge,
  • make Biodiversity data more accessible and visible in the world,
  • make the usage of KG easy for domain experts.

We suggest the following topics:

Creating and Maintaining KG

  • (Semi-)Automatic generation of Knowledge Graphs, tools& services
  • Domain-specific Knowledge Graphs creation
  • Subgraph creation
  • From Knowledge Graphs to graph embeddings & ontology creation
  • Ontology Learning in Biodiversity domain
  • Knowledge Graph with Citizen Science
  • Optimizing Knowledge Graphs
  • Maintenance of Knowledge Graphs
  • Biodiversity Knowledge Graphs using Machine Learning
  • Provenance tracking in the Knowledge Graph creation

Querying

  • Usage scenarios for Knowledge Graphs in the Biodiversity domain
  • Knowledge discovery on Knowledge Graphs
  • Leveraging Biodiversity Knowledge Graphs for Machine Learning
  • Knowledge Representation and Reasoning

Interacting

  • Interactive generation of Knowledge Graphs with non-semantic-web experts
  • Visualizing Knowledge Graphs
  • User interfaces (UIs) for Knowledge Graphs

Linking

  • Integration of existing Knowledge Graphs
  • Linking Biodiversity Knowledge Graphs
  • Linking paper, datasets, and code

Evaluating

  • Data Quality of Knowledge Graphs
  • Benchmarks

Session Chair : Samira Babalou (Institute for Computer Science, Friedrich Schiller University Jena, Germany)

Co-chairs :  Birgitta König Ries, Trupti Padiya, Felicitas Löffler Felicitas Löffler and Nora Abdelmageed

(Institute for Computer Science, Friedrich Schiller University Jena, Germany)

Machine and Deep Learning in Environmental Monitoring and Protection

To preserve our environment, several monitoring actions need to be considered in order to implement all policies and decisions to prevent and/or address any deterioration of natural habitat.
In recent years, Machine and Deep Learning (ML&DL) technologies have shown to be effective in a wide spectrum of environmental monitoring techniques for handling water, noise, air, and waste issues.
This session welcomes papers on all aspects of environmental monitoring and protection by means of novel and state-of-the-art ML&DL techniques in order to undertake environmental assessment work, interpret the results, and suggest appropriate corrective measures for sustainable development.
Session chair : Antonino Staiano (Università di Napoli Parthenope, Italy)
Co-chair : Ioannis N. Athanasiadisenope (Wageningen University and Research, Netherlands)

Digital Science for Environment

The session on “Digital Science for Environment” in ICEI 2020+1 event will mainly focus on research and developments in the area of application of cutting-edge technologies in the digital sciences along with mathematical solutions of the real-world problems in ecology in general.

The theme of the session are

  • Theoretical/mathematical approaches to ecological problems
  • Data Analytics/Big data applications in Ecology and Environmental Science specific to Environmental toxicity, Green Synthesis, renewable energy sources, sustainable agriculture suitable for environment protection
  • Infectious disease modelling

Session Chair : Manojkumar T K (Digital University Kerala, India)
Co-chair           : Radhakrishnan T, Ajithkumar R (Digital University Kerala, India)

Understanding and Improving Earth System Predictions with Emulator, Surrogate, and Hybrid Modeling

Predictions of the Earth system are limited by uncertainties in key ecological and ecohydrological processes–such as photosynthesis, transpiration, and decomposition–and uncertainties in relevant data products related to scale, heterogeneity, representativeness, and model structure. This session is focused on recent advances in machine learning that address such issues through (1) data-driven Earth system modeling, including hybrid process-/machine-learning parameterizations and surrogate-based modeling; (2) assessing the fidelity of ecological process representations in models through statistical and machine learning methods; and (3) developing, synthesizing, and evaluating data products of ecological, hydrological, and meteorological variables. Encouraged are contributions that focus on model benchmarking metrics, uncertainty quantification methods, and hierarchical modeling approaches employing novel model-data fusion and machine learning applications to improve the mechanistic understanding or representation of physical and ecohydrological responses and feedbacks to the Earth system, particularly under climate change.

Session chair : Forrest M. Hoffman (Oak Ridge National Laboratory, USA)
Co-chair : Jitendra Kumar (Oak Ridge National Laboratory, USA)

Earth Observation and Data Analytics for Ecological Monitoring

Earth, Ocean, Atmosphere, Planetary Science and Applications Area (EPSA) of Space Applications Centre (SAC) proposes to have a special session on “Remote Sensing for Ecological Monitoring”. As is well known, SAC is a center of Indian Space Research Organisation (ISRO) where sensors for Earth Observations (EO) are built and applications around them are developed. Sensors in Optical, Microwave, Hyperspectral and Thermal have already been developed and are providing systematic data from space platform. These are essential to study ecological and climate systems. EPSA has developed many societal Applications using vast and long time data sets. Prominent among them are related to Agriculture, Forestry, Coastal environment, Land degradation and Cryospheric studies (including Antarctic and Arctic regions).

EPSA has completed many national level projects in the thematic area of Agriculture, New and Renewable Energy, water resources, coastal Land use, desertification using EO Data as primary source. EPSA is engaged in developing archival and dissemination system for EO data and enriching VEDAS website (https://vedas.sac.gov.in) – a geoprocessing platform for visualization and rendering useful geo-spatial analysis on WEB – by adding new geo-spatial processing functionalities. Vast repository of EO Products from Indian and foreign satellites / sensors are hosted. Important recent geo-spatial information includes related to Vegetation Monitoring, New and Renewable energy, Hydrology and Urban sector.

 

The topic of the proposed special session is of interest to many institutes and organisations. Of late, World Bank is interested in having Data analytics output based geo-spatial outputs from long temporal EO data to monitor performance of the programmes supported by them. Prominent Institutes in India which have shown interest include National Remote Sensing Centre (NRSC) – Hyderabad, Indian Institute of Remote Sensing (IIRS) – Dehradun, North East Space Applications Centre (NESAC) – Shilong and Forest Survey Institute – Dehradun. More than 20 States have their RS / GIS Cells which may be interested in the topics to be covered in this special session. There are estimated 40 (forty) academic institutes in India which teach this subject at Master’s level and it is relevant for them as well.

SAC has demonstrated expertise in establishing and end – to – end utilization of space infrastructure for use of EO data for monitoring ecological processes. The cycle consists of defining sensor specification, developing sensors, data product generation, developing applications as well as tools for geo-spatial analysis on web using state of the art Data Science techniques and capacity building. Thus it is appropriate that EPSA – SAC utilises this very relavant platform to organize special session on “Earth Observation and Data Analytics for Ecological Monitoring” during ICEI – 2020.

Session Chair : Shashikant Sharma (Space Applications Centre (ISRO), India)

Co-chair : Bhattacharya B K (Space Applications Centre (ISRO), India)