Recommending Environmental Big Data Using Semantically Guided Machine Learning - Large Scale and Big Data: Processing and Management - page 490

Integrated knowledge RDF files were uploaded into the triple store, which pro-
vided all facilities of browsing, querying, exporting data in a different formats, etc.
RDF representation of the processed knowledge and PCA-SOM-based recommen-
dation made the knowledgebase easily interpretable to the data customers and to the
application designers. A user interface layer (Figure 15.16) was developed on top
of these triples to provide great flexibility to the application designers. A SPARQL
endpoint was available for navigating and downloading the already analyzed data
for greater usage.
15.5 CONCLUSION
Our understanding of the environment is greatly associated with the interlinked
knowledge of the phenomena surrounding us. Such knowledge is a result of data
and extracted information. With the availability of very high and even ultra-high-
resolution sensor data there is a greater need of managing data, information, and
essentially the knowledge. With the advent of technological novelties and their wider
applications, the generated data is surpassing our capacities to store it. There is an
urgent need for improved methods and advancement in data-intensive science to
retrieve, filter, integrate, and share data. Data and meaningful information are key
for the actors in every walk of life, however, how to conceive, perceive, recognize,
and interpret such data in space and time is a big question and a big challenge.
Taking this challenge into the perspective, we have presented an opportunity of rec-
ommending environmental Big Data using semantically guided machine learning
approaches. In our approach, we were successful in integrating various data sources
available in the web by developing web adaptors. For example, data from SILO,
COsMOZ, AWAP, ASRIS, and MODIS were processed within the machine learning
and image processing framework. The methodology resulted meaningful informa-
tion, which was later captured as knowledge. The knowledge was published in the
LOD cloud environment for the wider use of scientific community and policy-making
authorities. Currently, the most promising approach is LOD cloud in handling the
Big Data. By developing this methodology and recommending environmental big
knowledge, we believe that this approach will enable us to better address complex
scientific and social questions. We have a firm belief that our simple approach will
contribute to the body of knowledge in the Big Data study and big knowledge man-
agement in this era of data-intensive science.
REFERENCES
1. Don't Build a Database of Ruin. http://blogs.hbr.org/cs/2012/08/.
2. Frederic Achard, Guy Vaysseix, and Emmanuel Barillot. Xml, bioinformatics and data
integration. Bioinformatics , 17(2):115-125, 2001.
3. Sören Auer and Jens Lehmann. Creating knowledge out of interlinked data. Semantic
We b , 1(1):97-104, 2010.
4. Tim Berners-Lee. http://www.w3.org/DesignIssues/LinkedData.html/.
5. Muhammad Bilal and Sharifullah Khan. Ontology-driven relevance reasoning architec-
ture for data integration techniques. In Intelligent Systems, 2008. IS'08. 4th International
IEEE Conference , volume 3, pages 22-8. IEEE, 2008.