Information Technology Reference
In-Depth Information
Fig. 1. Plant cultivation LOD
a screen with photos of the cultivated plants. NEC Corp. offers a machine-
to-machine (M2M) service for visualizing sensor information and supporting
farming diaries. Both systems address recording and visualization of the work,
although our system is aimed at the search of cultivation knowledge on site by
means of a voice-controlled question-answering system.
Recently, Apple's Siri has drawn attention to the question-answering system.
Siri offers a high-accuracy voice recognition function and correctly answers the
question in the case that the information source is a well-structured website
such as Wikipedia. However, extracting the information from unstructured web
sites often fails and Siri returns the search engine results page, and then the
user needs to tap URLs from the list. Thus, LOD is a promising source of the
question-answering system. In fact, IBM's Watson uses Linked Data as the inter-
nal information source in part [ 4 , 5 ]. [ 6 ] serves as a useful reference for surveying
other question-answering systems, some of which uses LOD as the knowledge
source. Although our proposed system is related to a number of works, it is dis-
tinguished by accuracy improvement and data acquisition by the user feedback
and registration mechanism described in section 4 . Also, there is no similar work
in terms of agricultural application.
3 Open Data for Agricultural Information
3.1 Plant Cultivation LOD
Although the agroXML approach is ideal, adhering to a unique schema would
pose diculties for broader use. Therefore, we propose modeling cultivation
Search WWH ::




Custom Search