ABSTRACTThe research on this reviewpaper presents the complicated usage of prescribed drugs which perform in thezone of data mining for organizing large volume of data and using of complexfunction for performing more refined analysis using cloud platform. The aim ofthis paper is to understand the extensive and innovative frame that uses the socialmedia to characterize drug abuse. Therough idea of this survey is a analytical approach to analyze social media foracquiring the emerging trends in drug abuse by applying powerful techniquessuch as cloud computing and Map Reduce model. Thispaper describes how to capture important data to evaluate from networks like Twitter,Facebook, and Instagram. Big data techniques are used to mine the usefulcontent for analysis.
1. INTRODUCTIONSocialmedia is an internet based applications which can be used for sharinginformation, sharing creative ideasthrough a communication network channel. Currently, social media is used forenumerating the information regarding patients for understanding the symptomsof patient. Social media allows message sharing, collecting information anddeliver to the health care space. Health care space is the one that provide thedata’s of patient with their permission. The proper way of accessing data andprograms over the internet known as cloud. It model the social medias such asfacebook, twitter etc using network based analysis method.
Currently, the scientific researchoften requires vast amount of estimation during simulation and data processing.The scientific problem can be solved by automatic computational throughcollection or array list which is emerged by set of sensors. Themain aim of this paper is to use the social media as an informative source foranalyzing the illicit drug activities in the society. Data mining play animportant role in all stages during the development of drug. The use of datamining techniques during the drug development is mainly classified into twoareas:1.
New Effect of Drug Identification: conflict reaction occurs mostly, butsometimes new remedial effect occur and effects in some population.2.Suitableness in drug use.
Acrawler which basis in Map Reduce Model is performing the data mining task for thedistributed computation of data which is implemented in the framework ofHadoop. Data processing consists of three stages, first and second stages arecollecting information from different media sources and filter it which resultsin small dataset with data corresponding to solve the task. On the last stagethe small dataset which are analyzed using refined models. The main advantageof this paper is that to provide knowledge about the drug usage for a group ofpeople which are observed who rarely use drugs or not addicted to drugs andanother aim is to collect the reviews of patients which cause side effects dueto the drug and can prescribe another drug through media.Literature SurveyFromV. R.
Nagarajan, et at1 social media provide information for the field ofhealth informatics which includes Bioinformatics, Image informatics, Clinicalinformatics, Public health informatics etc. In this paper they use the methodscalled SOMS( an analysis to check the interrelationship between user posts and positive or negative comments on drugusage) and hierarchical clustering. This paper provide a framework whichevaluate the positive and negative symptoms of disease and also the sideeffectsof treatment common cancers lung cancers.FromJun Huan, et al2 frequent subgraph mining is an active research topic in thedata mining community. They use graph as a general model to represent the dataad can be used in several field like bioinformatics, web indexing, etc. Theproblem of frequent sub-graph mining is to find all frequent subgraphs from agraph database.
In this paper they propose a new algorithm FFSM(Fast FrequentSubgraph Mining) for the frequent sub-graph mining problem i.e., to reduce thenumber of redundant candidates proposed.FromMathew Herland, et al3 a bulk amountof data is produced within health informatics and analysis of this data is doneby big data techniques and big data allows potentially unlimited possibilitiesfor knowledge to be gained. This information can improve health care qualityoffered to patients. A several problem will arise while managing this bulkamount of data especially how to analyze data in a reliable manner.
This paperpresents big data tools and approaches for the analysis of health informaticsdata gathered at multiple levels including the molecular, tissue, patient andpopulation levels.FromDeepa Sharma, et al4 appearance of recent techniques for scientific knowledgecollection has resulted in large scale accumulation of information relatingvarious fields. Retrieval of data from huge knowledge base by typical queryways is an inadequate form. Therefore, cluster analysis is used for analysisand k means clustering algorithm is mostly used for data mining applications.
The analysis of the cancer data set with the k meanand then applying with the Som. This paperproposes a technique for creating knowledge retrieval more practical andefficient using som with K mean clustering technique, So as to get betterclustering with reduced quality.
