Research And Implementation Of Agriculture Blog Based On Hadoop

With the Internet technology rapidly developing, Internet resources have become a majorsource of information and the same holds true in the field of agriculture. However, unlikeother people, the farmers generally lack knowledge in using computers and it is difficult forthem to obtain the information they need from the Internet. With an ever-increasing amount ofinformation accumulated on the Internet, it is especially challenging to sort for the agriculturalinformation of potential interest to the farmers. Therefore, it is necessary to build aninteractive platform for them to share information. Blogging is one quite popular means ofinformation exchange on the Internet, but many current blogging websites are neitherspecialized nor kept most updated and hence not well suited for the farmers. To address theseissues, the work presented here studies and realizes a distributive, updated and extensiblefarming blog, on basis of the distributive storage and parallel computing. The main body ofthe research and the conclusions are summarized as the following:(1) The model build-up for agricultural blogs. Firstly, to meet the needs of the farmersfor agricultural blogs, we designed an integrated service system for agricultural blogs. Wemodified the current research mode on blogs accordingly, built a distributive storage modeland a parallel data processing model for agricultural blogs and effectively improved theclassic push-pull mode.(2) The design and realization of the agricultural blog system. In consideration of thefunctional needs, the service system and the basic model, we firstly outlined the overallframework of the agricultural blog system, designed the overall dataflow and the majorfunction modules of the system, and denoted the design of the major data processing steps(storage and access). Also, we carried out detailed design and denotation of the databasemodule. Finally, parallel data processing was realized by using MapReduce and detailedexplanation was given on how data process and access service in Coprocessor was realized.(3) Testing and data analysis. Firstly, we laid out the distributive test set and integratedMyEclipse development environment on basis of the framework of Hadoop. Subsequently, wetested the functions in the user interface of the farm blog system. Finally, we tested thefunctionality of the system using simulated large dataset. The testing result indicated thefeasibility and high efficacy of our farm blog system.