D. VISHNU VARDHAN RAJU1,

           1M.Tech,PG Scholar,Dept.of CSE,
Sri Venkateswara University College Of Engineering,

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

Venkateswara University,Tirupathi.                                

                             2Associate professor, Dept.of CSE,
Sri Venkateswara University College Of Engineering,

Sri Venkateswara University,Tirupathi.



selection is an important technique for data mining. Despite its importance,
most studies of feature selection are restricted to batch learning. Unlike
traditional batch learning methods, online learning represents a promising
family of efficient and scalable machine learning algorithms for large-scale
applications. The machine learning field has gained its momentum in almost any
domain of research and just recently has become a reliable tool in the medical
domain .Machine Learning (ML) is envisioned as a tool by which computer-based
systems can be integrated in the health care field in order to get more
efficient medical care. An ML-based methodology is described to build an application
that is capable of identifying disseminating health care information .It
extracts sentences from published medical papers that mention diseases and
treatments, and identifies semantic relations that exist between diseases and
treatments. In this paper we have focused on Machine Learning (ML) techniques
and the Classification algorithms that are suitable to use for identifying and
classifying relevant medical information in short texts. The encouraging
results of our experiments validate the efficiency of the proposed techniques.

Keywords:- Bigdata Analytics, Machine Learning
, Health Care. 

                               I.  INTRODUCTION

 People care deeply about their health and want
to be, now more than ever, in charge of their health and healthcare. Life is
more hectic than has ever been, the medicine that is practiced today is an
Evidence-Based Medicine (hereafter, EBM) in which medical expertise is not only
based on years of practice but on the latest discoveries as well. Tools that can
help us manage and better keep track of our health such as Google Health and
Microsoft HealthVault are reasons and facts that make people more powerful when
it comes to healthcare knowledge and management. The traditional healthcare
system is also becoming one that embraces the Internet and the electronic
world. Electronic Health Records (hereafter, EHR) are becoming the standard in
the healthcare domain. Researches and studies show that the potential benefits
of having an EHR system are:


Health information
recording and clinical data repositories: immediate access to patient
diagnoses, allergies, and lab test results that enable better and
time-efficient medical decisions.


rapid access to information regarding potential adverse drug reactions,
immunizations, supplies, etc.


the ability to capture and use quality medical data for decisions in the
workflow of healthcare.


treatments that are tailored to specific health needs: rapid access to information that is
focused on certain topics.


In order to embrace the views that
the EHR system has, we need better, faster, and more reliable access to
information. In the medical domain, the richest and most used source of
information is Medline, a database of extensive life science published
articles. All research discoveries come and enter the repository at high rate,
making the process of identifying and disseminating reliable information a very
difficult task. The work presented here is focused on two tasks: automatically
identifying sentences published in medical abstracts (Medline) as containing or
not information about diseases and treatments, and automatically identifying
semantic relations that exist between diseases and treatments, as expressed in
these texts. The second task is focused on three semantic relations: Cure,
Prevent, and Side Effect.


The tasks that are addressed here are the foundation
of an information technology framework that identifies and disseminates
healthcare information. People want fast access to reliable information and in
a manner that is suitable to their habits and workflow. Medical care related
information (e.g., published articles, clinical trials, news, etc.) is a source
of power for both healthcare providers and laypeople. Studies reveal that
people are searching the web and read medical related information in order to
be informed about their health. The proposed objective for this work is to show
the different Natural Language Processing (NLP) and Machine Learning (ML)
techniques are used for the representation of information and  classification algorithms that are suitable to
use for identifying and classifying relevant medical information in short

                        II. RELATED WORK

data set consists of sentences from Medline abstracts annotated with disease
and treatment entities and with eight semantic relations between diseases and
treatments. All the way the information is provided as abstracts in form of the
documents. It is possible that the abstracts are made largely available to the
viewers,but it is inconvenient to measure the ratio of facts , our research is
focused on different representation techniques, different classification
models, and most importantly generates improved results with less annotated

existing studies of online learning require accessing all the
attributes/features of training instances. Such a classical setting is not
always appropriate for real-world applications when data instances are of high
dimensionality or it is expensive to acquire the full set of
attributes/features. To address this limitation, we investigate the problem of
Online Feature Selection (OFS) in which an online learner is only allowed to
maintain a classifier involved only a small and fixed number of features. The
key challenge of Online Feature Selection is how to make accurate prediction
for an instance using a small number of active features. This is
in contrast to the classical setup of online learning where all the features
can be used for prediction.

1) Prediction of heart disease at early stage using data mining and big
data analytics: A survey                      

In this paper, the various technologies of data mining (DM) models for
forecast of heart disease are discussed. Data mining plays an important role in
building an intelligent model for medical systems to detect heart disease (HD)
using data sets of the patients, which involves risk factor associated with
heart disease. Medical practitioners can help the patients by predicting the
heart disease before occurring. The large data available from medical diagnosis
is analyzed by using data mining tools and useful information known as
knowledge is extracted. Mining is a method of exploring massive sets of data to
take out patterns which are hidden and previously unknown relationships and
knowledge detection to help the better understanding of medical data to prevent
heart disease. There are many DM techniques available namely Classification
techniques involving Naive Byes (NB), Decision tree (DT), Neural network (NN),
Genetic algorithm (GA), Artificial intelligence (AI) and Clustering algorithms
like K-NN, and Support vector machine (SVM). Several studies have been carried
out for developing prediction model using individual technique and also by
combining two or more techniques. This paper provides a quick and easy review
and understanding of available prediction models using data mining from 2004 to
2016. The comparison shows the accuracy level of each model given by different

2) The eICU Research Institute –
Collaboration between Industry, Health-Care Providers, and Academia

In this paper as the volume of data that is electronically available
proliferates, the health-care industry is identifying better ways to use this
data for patient care. Ideally, these data are collected in real time, can
support point-of-care clinical decisions, and, by providing instantaneous
quality metrics, can create the opportunities to improve clinical practice as
the patient is being cared for. The business-world technology supporting these
activities is referred to as business intelligence, which offers competitive
advantage, increased quality, and operational efficiencies. The health-care
industry is plagued by many challenges that have made it a latecomer to
business intelligence and data-mining technology, including delayed adoption of
electronic medical records, poor integration between information systems, a
lack of uniform technical standards, poor interoperability between complex
devices, and the mandate to rigorously protect patient privacy. Efforts at
developing a health care equivalent of business intelligence (which we will
refer to as clinical intelligence) remains in its infancy. Until basic
technology infrastructure and mature clinical applications are developed and
implemented throughout the health-care system, data aggregation and
interpretation cannot effectively progress. The need for this approach in
health care is undisputed. As regional and national health information networks
emerge, we need to develop cost-effective systems that reduce time and effort
spent documenting health-care data while increasing the application of
knowledge derived from that data.

                              III. METHODOLOGY

In this paper
focus is on different representation techniques, different classification
models, and most importantly generates improved results with less annotated
data. The tasks addressed in our research are information extraction and
relation extraction. From the wealth of research in these domains, There are
three major approaches used in extracting relations between entities
Co-occurrences analysis, Rule based approaches, Statistical methods. Statistical methods tend to be used
to solve various NLP tasks when annotated corpora are available. Rules are
automatically extracted by the learning algorithm when using statistical
approaches to solve various tasks. In general, statistical techniques can
perform well even with little training data. For extracting relations, the
rules are used to determine if a textual input contains a relation or not.
Taking a statistical approach to solve the relation extraction problem from
abstracts, the most used representation technique is bag-of-words

two tasks that are undertaken in this paper provide the basis for the design of
an information technology framework that is capable to identify and disseminate
healthcare information. The first task identifies and extracts informative
sentences on diseases and treatments topics, while the second one performs a
finer grained classification of these sentences according to the semantic
relations that exists between diseases and treatments.




                                         Fig 3.1 Architecture



3.1 shows the details of the online performance of the compared online feature selection
algorithms with varied fractions of selected features. The proposed OFS
algorithms outperform the other two baselines for most cases.

encouraging result further verifies the efficiency of the proposed technique
Application Modules.




1.Collect abstracts




3.Text extraction


4.2c – ct—                                                                                                                                                                                     



9.goto step 3

 10.End if

 11.Update Training Set Ts