Towards an effective approach for cyanobacteria affected locations extraction from news feeds
Jadhav, Anuja C
MetadataShow full item record
Blue-Green Algae (BGA) are a toxic phytoplankton which are now a worldwide phenomena and a concern to various public authorities. The issue of cyanobacteria was brought to a highlight with its first ever official published report in 1878. The thick belts of these BGA (also known as Cyanobacteria) restricts the sunlight from penetrating into the water bodies which leads to depletion of the levels of dissolved oxygen thus hampering the aquatic life as well as affecting the humans coming in contact of the contaminated water. The year 2016 has seen extensive algal bloom for some prominent and large water bodies like Lake Okeechobee. With our work we aim to extract information of locations affected by cyanobacteria from news articles. We first illustrate the purpose of analyzing news articles with respect to BGA and, then formulate this purpose into an extraction workflow using unsupervised machine learning and named entity recognition framework based data mining approach. We evaluate our approach by validations from supervised learning techniques on data set gathered from news articles for the year 2017. Our experiment shows that our approach is effective for the locations detection problem statement in the cyanobacterial news articles domain.