Find a scholarly article that includes empirical research. In other words, the a

Question

Find a scholarly article that includes empirical research. In other words, the article should include some sort of quantitative, qualitative, simulation, or experimentation as a form of research and should include findings from the data collection. Write a paper that addresses the following:

1. Describe the motivation for the study. What problem(s) do the authors address?

2. Briefly, summarize the authors’ findings and contribution.

3. Discuss why the study significant. Who is affected by the problem addressed and the study’s findings? Justify.

4. Discuss a modern scenario in which this study is relevant. Describe the environments in which the scenario may occur.

Support your paper with a minimum of two additional external resources. In addition to these specified resources, other appropriate scholarly resources, including older articles, may be included.

Explanation / Answer

paper title : Disease Inference from Health Related Questions via Sparse Deep Learning

type : IEEE

1. Motivation :

The final goal of automated health system is disease inference. According to a user study
of health automated system that the health seekers frequently ask for :
1. How to overcome with the problems they are having.
2. How to prevent the problems inferred from the symptoms they are having.
3. Possible disease types they are having inferred from the symptoms.

2.Contribution to the project :

To automate the answers that are provided by authorised doctors on the websites for health related problems using sparsed deep learning neural networks on the signatures that are mined from the question and answer data set collected from the healthtap api and other health blogging websites using beautifulsoup api The objectives for automated disease inference are the following:
• The model must be able to mine the signature words from the questions or queries that are provided by he users on th websites.
• The proposed system is expected to overcome the lack of proper vocabulary in the user’s query that fails to provide the doctors accurate medical term for the illness.
• to develop a sparsed neural network with the first layer taking the raw input features
from the QA datasets and three hidden layer to get the signatures using the graphs that are built over the terms used in the raw input features.
• The above structure of one hidden layer of signature is considered as input features for the other three hidden layer which uses the sigmoid function for pretraining and for further optimization we have also proposed to autotune the layers.

3.Who will affect ?

The designed system , automates where it automates the questions that are related to the users.That means we had a website (related to health) webMd which takes questions (user queries)from the user ,for that question health experts will give the answers .We have tried to automate this process.That means whenver a new user comes the answer has to generate from existing data.

4. Modern scenario

This paper first performed user study to analyze the health seeker needs. This provides the insights of community based health services. It then presented a sparsely connected deep learning scheme that is able to infer the possible diseases given the questions of health seekers. This scheme is constructed via alternative signature mining and pre-training in an incremental way. It permits unsupervised feature learning from other wide range of disease types. Therefore, it is generalized and scalable as compared to previous disease inference using shallow learning approaches, which are usually trained on hospital generated patient records with structured fields.We hae developed a module that can translate the user’s vocabulary to normalized medical terms which has wide range of applications when we are
building medical bots and AI structuring in a robot which could prove to be an important module given future of robots and AI bots in medical field.The model developed in this work is to our best knowledge a first attempt over the blog data , apart from just medical inference it could applied over other data to obtain a better language modelling when transitivity from user language to a standard vocabulary is needed.A responsive system wherein further response could be asked to the user after inferring on primary query to get better and accurate results like, for example when i give symptoms for a disease like migraine it can further ask me whether symptoms related to migraine are faced by me which could also include a module on image based classification wherein the user may be wich could also include a module on image based classification wherein the user may be prompted to either take a picture of affected part in order to give a better validated results, which could be future work associated with this project work.

References :

[1] S. Doan and H. Xu, “Recognizing medication related entities in hospital discharge summaries using support vector machine,” in Proc. Int. Conf. Comput. Linguistics,2010, pp. 259–266.
[2] T. D. Wang, C. Plaisant, A. J. Quinn, R. Stanchak, S. Murphy, and B. Shneiderman, “Aligning temporal data by sentinel events: Discovering patterns in electronic health records,” in Proc. SIGCHI Conf. Human Factors Comput. Syst., 2008, pp. 457–466.
[3] M. Shouman, T. Turner, and R. Stocker, “Using decision tree for diagnosing heart disease patients,” in Proc. 9th Australasian Data Mining Conf., 2011, pp. 23–30.

Related Questions

Navigate

• Browse (All)
• Browse F
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.