Automated agricultural ontology Construction using machine Learning architecture

Abstract

Evolution of domain specific ontology and its deployment is newlinethe latest fad in the field of Knowledge Representation (KR). These newlinedays, several applications are big data oriented with the stipulation newlineof using excellent data processing mechanisms, one such drive is newlineautomatic creation of ontology. These sort of ontologies can easily newlinehandle domain specific query. Considering which, in this research we newlinebring forward the depiction of an agricultural ontology. The field newlineof agriculture itself is very vast wherein information is handy as in text newlinefiles, tables and spreadsheets formats, often underutilised due to the newlineinadequacy of contemporary informatics deployment. Thus, in this newlineresearch we intend to develop an intelligent approach for ontology newlineconstruction with automatic alignment of relation instances for newlineunstructured texts in the Knowledge Base (KB). The process begins newlinewith extracting relational words from input and using them for newlineconstructing KB with Word Net. newlineEvery design methodology involves data processing, likewise newlinein our case Natural Language Processing (NLP) techniques are newlineemployed for tagging the terms, which further provides dimension newlinereduction to form Formal Concept Analysis (FCA) and ensuing newlinedepiction. The sub ontologies built on fundamental factors e.g. weather, newlinesoil, and pests are gathered to develop a comprehensive ontology. The newlineontology construction for climate factors has been designed using newlineHybrid Neural Network for Agricultural Ontology Construction (HNNAGOC) newlineand further classification is carried out using Recursive Neural newlinex newlineNetwork (RNN) to forecast effective methods towards creating a robust newlineontology. Furthermore, the designed quotAutobotologyquot explores necessary newlinedatabase from DBpedia Virtuoso Server, to respond user s query over newlineVirtuoso Faceted agriculture- based web services with unique web newlineexperiences for user. newlineThis research has been performed over Domain - Specific newlineLong-Short Term Memory Model (DS- LSTMM) recommended for newlineagricultural informatics. Finally, an innovative Naïve Bayes Organizer newlinealong