Social Media Networks in Online Health Care for Topic Analysis And Sentiment Analysis Using Text Mining Techniques

Transcription

1 Volume 118 No , ISSN: (printed version); ISSN: (on-line version) url: ijpam.eu Social Media Networks in Online Health Care for Topic Analysis And Sentiment Analysis Using Text Mining Techniques M.Surya prabha 1 Dr.B.Sarojini 2, 1 Research Scholar, 2 Assistant Professor, Department of Computer Science, Department of Computer Science, Avinashilingam University, Avinashilingam University, Coimbatore-43, India, Coimbatore-43, India, surya.anju05@gmail.com dr.b.sarojini@gmail.com Abstract --- Social media networks are connecting 0patients, health seekers in online communities to improve the patients health care, and also facilitate communication between medical professionals and health seekers to ask health related queries, share information s, opinions, observations. There are number of social media networks and applications have been specifically created for doctors and other care givers in health care organizations. Online health care Topic Analysis are useful for patients, caregivers and different stack holders are interested to analyse and read the topics for their disease related symptoms and get better solutions for their health issues. This paper aimed to specify content and topic analysis using in text mining techniques applied for online health care social media. Data mining and machine learning algorithms are used to analyze the data in the online health care applications. Keywords: Social media networks, Online Health Care Topic Analysis, Sentiment Analysis, Text mining. I. INTRODUCTION Social media platforms facilitate information creation, sharing informations to others and retrieval, creating new links for reaching patients or other online health care consumer s. It gives a better solution and immediate information sharing and patients seem to be way ahead of healthcare professionals when using digital technology to access health information [1]. online Health care Information and communication technologies are used for transform how health seekers and patients think about access and digitally provide health care information [2].There are some advantages using social media technologies for patients, health professionals,online health experts and care givers. Some reasons why Patients and care givers are involved in social media health care is, they get some knowledge about the health issues and disease related problems from peer who are affect same health issues. Online health care Communities are very helpful for patients, care givers and other community users in the social media networks they feel health experts and doctors are very helpful for asking questions and answers for their health related problems [3]. Some patients are willing to provide and share their medical experiences with treatment are considered valuable Information for the patients with newly diagnosed. 2929

2 These health seekers help other patients to get answers in their time of need. These health information providers are recognized as health experts and make the contributions to health care social media. Online health care social media is also provides a platform for Emotional support and communication for users, health professionals and patients. Some emotional expressions such as conveying, communicate some positive thoughts and Emotional talks are better solution for patients who get some solutions for their health issues [4]. It is importantly, this study work is very help to the users, Patients, health seekers in online health care communities; especially the beginners, caregivers and patients. Beginners and health seekers are getting difficult to directly understand this new process of online health care communication, so online health care topic analysis and sentiment analysis are obtain, what are the needs in online health care communities and find health issues quickly accessing are become very useful involvement in online health communities more easily, here by getting valuable information for their healthcare issues and queries. For these reasons, there are many research works are involved in health related topics in online health Care communities. Some of the research studies and methodsare based on questionnaires and statistical content analysis. In previous studies, health care information are shared in social media online health care communities were determined according to the number of people, patients and users who used the list and how frequently they posted comments and messages on online health care community [5]. Previous online health care related topic study basedon text mining of medical text isprimarily focused on clinical decisions and medical literature. In recent years, however, person generated scientific text has been shared on many social media offerings, inclusive of clinical weblogs, healthq&a, and online health groups. Denecke (2009) targeted on clinical weblogs and categorized the topics in clinical weblogs into two methods, informative and affective. Brody (2010) used text classification based on latent Dirichlet allocation (LDA) in topic models to detect the relevant components of online reviews of health care experts. Chen (2011) accomplished a cluster techniqueson clinical posts from three online health groups and found that the clusters might be categorized into a fixed of not unusual classes, basic, support, patient-focused, discovered understanding, remedies or methods, treatments and condition management. Fig 1 Social Media in Online health care In this research work, clustering techniques are involvedvaluable for assuming out different types of topic facts. In online health care communities there is large quantity of posts and comments generated by using customers. This data are unstructured, a few studies analyze the content material of fitness subjects and information from the internet. There s a need of textual content mining strategies. It is valuable and analyzes the statistics and mined the retrieval records from vast amount of records inside the online health care forums [6]. II. LITERATURE STUDY Social media analysis in online healthcare recognized different effects and health analysis use by patients for health related issues and reasons within the online healthcare system. Social media can support to patients. For example, it adopts their self-support by complementing the information provided by healthcare professionals [7]. These days the arrival of social media contributions involving of Health care Wikipedia, twitter, online forums and message forums, in which health seekers are more anticipated to manage health care information and share health studies on those social media and health care websites. Most recent survey established that 80% of net users have searched online healthcare problems, such as certain disease or treatment, 34% of them have study a another person statement or revel in about disease issues or medical problems in online health group, internet site, or weblog, and 24% of them have consulted online assessments of detailed medicines or clinical therapies [8]. Social media networks and applications use with the aid of patients does not most effective 2930

3 provide favorable results for health associated. It may also establish a challenge within the healthcare system to both patients and healthcare professionals. In online health care users can post health care information on how to deal with a certain health condition, it is essential to create reliable online communication to prevent health problems [9]. The vast amount of online health related messages posted by social media users, so it is very difficult and time intense to employ traditional statistical approaches to find valuable health care information. Moreover, social media consumers are generally unwilling to suggest personal information due to privacy concerns, and user data are usually registered and hence it is not easily accessible. To tackle these complications, text mining techniques are recently applied in online health care social media in some research studies [10]. Some research works are using text mining techniques to this user generated medical contents to explore the health care topics that concern online health care information users in online health care forums. III. ONLINE HEALTH CARE TOPIC ANALYSIS AND BENEFITS In Recent year s patients, health seekers and care givers are mostly involved in social media and online health care communities. Hot topics and health problem topics from online communities are helpful and better understand for health related Knowledge. There are some reasons patients and care givers are more involved in internet and online communities, patients feel doctors are very busy to explain disease related and health related topics for patients and health seekers. Many doctors explain only basic medical related answers for their questions. Internet and online communities are more effective for patients and care givers to get health decisions and make treatment benefits from peers who are in social media and other online communities. Health Care Topic analysis strategy have been broadly used to process in medical textual content. Previously topic analysis based on text mining of medical text was targeted on medical descriptions and medical texts. In recent times user generated medical textual content has been collective on various social media sites, scientific weblogs, healthq&a, and online health care groups. A few research work implemented in text mining techniques to this user generated clinical text to discover the topics that interest on knowledge online health data. A. Text Mining Techniques and Applications Text mining techniques refers from textual data. Text Mining (TM) Techniques are increased in recent years due to the enormous amount of text data, which are created in a Variety of forms such as social networks, patient records, health Care insurance data, news outlets, etc. The process of mining high quality of information from text is structured, semi-structured and unstructured text resources such as word documents, videos, and images. It generally covers a large set of related topics and algorithms for analyzing text, covering various groups, including data retrieval, Natural language processing, and Data mining, Machine learning and many application domains. Some of the text mining approaches are given here. B. Natural Language Processing (NLP) NLP is one of the text analysis methods in Text Mining techniques. It analysis the text formats which are understand by the machine read text. It tends to focus the text into word phrasing, Word Stemming (Removing suffixes), POS tagging (noun, verb, preposition etc.) in the word sentence, multi phrase removal in text etc. C. Information Extraction Information Extraction is to extract the information from structured, unstructured and Semi structured data in text or documents. Information Extraction tasks are source selection, Named Entity reorganization, Tokenization Normalization, Instance Extraction in the Text. D. Preprocessing Methods in Text Mining Preprocessing methods are one of the most important techniques in text mining. Preprocessing step usually consists of some following tasks such as text tokenization, stemming, filtering lemmatization etc. It was briefly described here as E. Text Tokenization Tokenization: Tokenization is the task of splitting a character categorization into (words/phrases) called tokens, and perhaps at the same time throws away certain characters such as punctuation marks. 2931

4 Step1 Step2 Data collection from Online Health Care Community forums G. Filtering Filtering method is used to remove some words in the given text or documents. A familiar filtering method is stop word removal in stop word removal method Words are frequently applied in the text without having much content information. Example prepositions, conjunctions etc. IV. SENTIMENT ANALYSIS Text Tokenization Text Preprocessing Stemming Filtering Sentiment analysis also defined as opinion mining and emotional analysis. In social media networks, and online health care communities, and online shopping administrations are developed various information and they need users and consumer reviews, opinions, and rankings for their websites.sentiment analysis in healthcare practices natural language software to categorize and consider written and spoken comments by patients about their healthcare experience. Step3 Step4 Apply Text Mining Techniques and Algorithms A massive amount of social media websites contains sentences that are concerned in sentiment-based. Sentiment analysis involves the use of Natural Language Processing (NLP), Data, or machine learning methods to mine, classify, or describe the sentiment content of a text source. A. Sentiment Analysis in Text Mining Step5 Fig 2 Frame work for topic analysis from online health care using Text mining F. Stemming Sentiment Analysis Topic Analysis from online Health care Sentiment analysis also known as opinion mining and subjectivity analysis. This process is to determine the polarity of opinions or reviews written by humans to rate products or services. It Defines if a word expression is Positive word, Negative word, or Neutral, and to what degree. In other words, text analytics studies the expression value of the words, with the sentence structure and the relationships among the words. Text analytics gives the meaning of the words. Sentiment analysis determines emotion words in the given text.sentiment Extraction retrieves the opinions or emotion words in the unstructured text to be analyzed. In this sentiment text extraction rating of words such as adjectives or adverbs that involve a key role in determining polarity of a sentence. Example positive words are good, happy, excellent etc. words that are negatives are bad, poor, disappoint, indelible etc. These are the text analysis in the sentiment score and extraction in sentiment analysis. Stemming is used to find the purpose in stem (root) of the derived words in the text or sentence. For example if the word connections, connectivity, connection, is to be stem by the Connect. It is used for reduce the suffix words in the given text. 2932

5 Text from datasets Fig 3 Framework for sentiment analysis in Text V. CONCLUSION In Online health care communities health-related topics, disease related symptoms, and health issues are important for website providers, information researchers, patients, and care givers. In this paper, text mining techniques are used to explore healthrelated hot topics and sentiment based text analysis in online health care communities. Recently text mining techniques used in online health related topic discussions and sentiment analysis in previous studies. By analyzing the health related topics and disease related topics are useful for patients, health seekers and users in the online health care community. This valuable online health information provides a better understanding and solutions for health care social media use by patients, Health seekers and care givers in online health care community. REFERENCES TokenizedSentence Sentiment Extraction Sentiment Score Results support groups. Patient Education and Counseling 87: , (2011). [7] Macias W, Lewis LS, Smith TL. Health-related message boards/chat rooms on the Web: discussion content and implications for pharmaceutical sponsorships. J Health Commun10(3): [Doi: / ] [Medline: ], 2005 Apr. [8] Fox S, Jones S The social life of health information. Pew Internet. (2009). [9] Availableat: Lifeof-Health-Information.aspx. [10] Rupert DJ, Moultrie RR, Read JG, Amoozegar JB, Bornkessel AS, Donoghue AC, Sullivan HW. Perceived healthcare provider reactions to patient and caregiver use of online health communities. Patient EducCouns. 96(3):320 6, [11] Brody S, Elhadad N Detecting salient aspects in online reviews of health, (2010). [12] Providers. AMIA Annual Symposium , [13] Carter M. Medicine and the media: How Twitter may have helped Nigeria contain Ebola. Br Med J. 2014; 349. [14] Sadah SA, Shahbazi M, Wiley MT, Hristidis V. A study of the demographics of Web-based health-related social mediausers. J Med Internet Res 17(8):e194 [FREE Full text] [doi: /jmir.4308] [Medline: ], [15] Fox S, Jones S The social life of health information. Pew Internet. Available at: The-Social-Lifeof-Health-Information.aspx (2009). [16] AndreasHotho, Andreas Nürnberger, and Gerhard Paaß. A Brief Survey of Text Mining. In Ldv Forum, Vol [17] Anne Kao and Stephen R Poteet Natural language processing and text mining. Springer. [18] K Bretonnel Cohen and Lawrence Hunter Getting started in text mining. PLoS computational biology 4, 1 (2008). [19] Sophia Ananiadou, SampoPyysalo, Jun ichitsujii, and Douglas B Kell Event extraction for systems biology by text mining the literature. Trends inbiotechnology 28, 7, , (2010). [20] Christopher D Manning, PrabhakarRaghavan, and HinrichSchütze. Introduction to information retrieval. Vol. 1. Cambridge University press Cambridge, [21] Christopher D Manning, HinrichSchütze, et al Foundations of statistical natural language processing.vol MIT Press. [22] Prithviraj Sen. Collective context-aware topic models for entity disambiguation. In Proceedings of the 21st international conference on World Wide Web.ACM, , [23] Li N, Wu DD, Using text mining and sentiment analysis for online forums hotspot detection and forecast. Decision Support Systems 48: , (2010). [1] Adams, S. Revisiting the online health information reliability debate in the wake of web 2.0 : [2] An inter-disciplinary literature and website review. Int. J. Med. Inform., 79, [CrossRef][PubMed], [3] Reis, S.; Visser, A.; Frankel, R. Health information and communication technology in healthcare communication: The good, the bad, and the transformative. Patient Educ. Couns. 93, [CrossRef] [PubMed],2013. [4] Sorana Daniela Bolboaca 1,* and Cristina Drugan2, Social Media Usage for Patients and Healthcare Consumers: A Literature ReviewAriana-Anamaria Cordo s 1, [2017]. [5] Lu Y, Zhang P, Liu J, Li J, Deng S (2013) Health-Related Hot Topic Detection in Online Communities Using Text Clustering. PLoS ONE 8(2): e56221.doi: /journal.pone [6] Chen AT Exploring online support spaces: Using cluster analysis to examine breast cancer, diabetes and fibromyalgia 2933

6 2934