Adequacy of Personal Medical Profiles Data in Medical Information Decision-Making Support System Solomiia Fedushko [0000-0001-7548-5856] Lviv Polytechnic National University, Lviv, Ukraine solomiia.s.fedushko@lpnu.ua Abstract. The paper describes the verifying methods of medical specialty from user profile of online community for health-related advices. To avoid critical situations with the proliferation of unverified and inaccurate information in medical online community, it is necessary to develop a comprehensive software solution for verifying the user medical specialty of online community for health-related advices. The algorithm for forming the information profile of a medical online community user is designed. The scheme systems of formation of indicators of user specialization in the profession based on a training sample is presented. The method of forming the user information profile of online community for health-related advices by computer-linguistic analysis of the in- formation content is suggested. The system of indicators based on a training sample of users in medical online communities is formed. The matrix of medi- cal specialties indicators and method of determining weight coefficients these indicators is investigated. The proposed method of verifying the medical spe- cialty from user profile is tested in online medical community. Keywords: Personal Medical Profiles, Decision-Making Support System, Med- ical Information, Profiles Data, Health-Related Advices. 1 Research Aims  developing community health workers  improving quality of e-health systems  simplifying the patients identification in various medical organizations  collecting, processing and analyzing Big medical data from various sources  working-time reduction of medical staff  detecting non-valid accounts and accounts with incorrect or stale medical data  consolidating all medical personal data with high level adequacy Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons Li- cense Attribution 4.0 International (CC BY 4.0) IREHI 2018 : 2nd IEEE International Rural and Elderly Health Informatics Conference 2 2 Problems Solving To develop an effective methods and means of determining the level of data ade- quacy of personal medical profiles. Fig. 1. E-health systems categories 3 Models of personal medical profiles Proposed method is carried out in order to identify the user personal data in the online community and verify the authenticity of the personal data specified by the user. The information track of the web user: InfTrack  Pi   Content  Pi  , PersonalData  Pi  (1) The components of an information track are: Content(Pi) created by a member of the online community, and personal data – PersonalData(Pi). Content  Pi   Thread  Pi  , Poll  Pi  , Post  Pi  (2)   j 1 NiUThead Thread  Pi   Thread j  Pi  is set of online community discussions;   j 1 is set of online community polls; NiUPoll Poll  Pi   Poll j  Pi  3   j 1 is set of online community posts. NiUPost Post  Pi   Post j  Pi  Computer-linguistic analysis is made only for the personal data that the user of the online community has specified in user account. The most prioritized data for forming the data profile of the patients in the online community is the mandatory information about the online community user, less im- portant – important data. Personal data distribution of online community user account to blocks is as follows: BasicInfo  Pi  , EduInfo  Pi  , InterestsInfo  Pi  , PersonalData  Pi   (3) WorkInfo  Pi  , ContactInfo  Pi  , FotoInfo  Pi  , Formal description of the online community member account: BasicInfo(Pi) is block of basic personal information of the online community user. Name  Pi  ,NickName  Pi  ,Age  Pi  , BasicInfo  Pi = (4) Gender  Pi  , Region  Pi  ,Lang  Pi  where Name(Pi) is full name; NickName(Pi) is nick name; Gender(Pi)is gender; Age(Pi) is age; Lang  P   Lang  P Ni is plural of languages signed by a user; Lang i j i j 1 Region  Pi =  Regionk  Pi  i N Region NiLang is number of language; is set of regions with k=1 which a user is associated; NiRegion is set of regions. EduInfo(Pi) is information block about education. EduInfo  Pi   EduLevel  Pi  , Specialization  Pi  (5) where EduLevel(Pi) is level of education received, Specialization(Pi) is specialty. WorkInfo(Pi) is block of data about the work of online community user. WorkInfo  Pi   Company  Pi  , Position  Pi  (6) where Company(Pi ) is institution where works, Position(Pi) is a position taken by a member of the online community in that institution. ContactInfo(Pi) is contact information for the member of the online community. ContacInfo  Pi   Email  Pi  , SocialNets  Pi  ,Website (7) where Email(Pi) is main email address, SocialNets  P   SocialNets  P Ni (Up ) is i j i j 1 plurality of pages in social networks, N iUp  is the number of pages in social networks; Website is website. FotoInfo(Pi) is graphic information block. 4 FotoInfo  Pi   Avatar  Pi  ,Userbar  Pi  , Foto  Pi  (8) where Avatar  P    Avatar  P Ni Ava is set of avatars, N iAva is number of avatars, i j i j 1 N iFoto is a number of photos, Userbar  Pi   Userbark  Pi  i Userbar N is set of graphic k 1 signatures, NiUserbar is number of signatures, Foto  Pi   Fotom  Pi  Ni Foto is set of photos. m 1 InterestsInfo(Pi) is a block of information about the hobby and interests of the online community user. InerestsInfo  Pi   Byline  Pi  , Activity  Pi  , Quot  Pi  , Biography  Pi  (9) where Byline  P   Byline  P Ni Byline Byline is number of signatures, N i is number of i j i j 1 signatures; Activity  Pi    Activityk  Pi  Ni Act is a set of favorite lessons and phrases, k 1 N iAct is number of lessons, phrases, Quot  Pi   Quotl  Pi  Ni Quot is a plurality of l 1 quotations, N iQuot is number of quotations, Biography(Pi) is a biography. Information about contacts and websites where the web user displays communica- tive activity is placed in the ContactInfo(Pi) block. Each account blocks contain in- formation from three groups of personal data of the online user. Preferably, in the BasicInfo(Pi) block, compulsory data is placed, without this data registration in the online community is not possible. 4 Method of determining the personal medical profiles data adequacy level The concept of the personal data adequacy of the patient profile of the medical clinics is presented to compare the personal data of the online communities informa- tional profile with the medical information system data. The information profile in the web communities is engendered by the method of computer-linguistic analysis of the user information track. Determining the personal data adequacy of the account to the real information of system user consists in the implementation of the main stages of the algorithm of determining the of personal data adequacy of the account. 5 Fig. 2. Block diagram of the algorithm for determining the personal data adequacy of account The difference between 1 and  jk  Value, P  is the distance between the reference value of the personal characteristics and the value of the personal data of the atomic k-th user is determined by the adequacy of the personal data of the k-th user profile.  j  Value, P   1   j  Value, P  k k (10)  j  Value, P  is distance to each possible value of the personal data of the atom- k ic k-th user of web community:   *w N_Ind  PrCh,k   PrCh,Vc   PrCh,P  2 μj  Value,P =1- k i=1 Indi,j -Indi,j  PrCh  i (11) where k  1 N _Vl  Pr Ch,Vc  . Moreover,  jk  Value, P   0,1 .  j  Value, P   max , then the degree of probability of personal characteristics of a k particular user in the online community to this user personal data is high. The proposed method of vectorization consists in transforming the data into a vec- tor form, which will enable to determine the degree of similarity between the values of personal characteristics. The value of a similarity measure between the value of personal characteristics and the control vector indicates the importance of member- ship by the online community to a certain value characteristics. 6 5 Approbation of results in the practice The results of data level adequacy analysis of personal medical profiles of Ukraini- an medical centers comparing to patients online information tracks. Fig. 3. Results of medical profiles data adequacy analysis of Ukrainian medical information systems The indicator of the effectiveness of the developed methods of data verification of personal medical profiles is determined in equation (12). N VerPD , NVerPD ¹N  LAdequacy  APD (12) Efficiency= -N  LAdequacy  VerPD APD N N(LAdequacy)APD is number of personal medical profiles with low account data adequacy, NVerPD is the total number of verified personal medical profiles. Based on equation (12) the results of data verification level of personal medical profiles, investigated profiles are classified according of data verification of personal medical profiles (21% of all investigated accounts contained high level of data verifi- cation, 41% of all investigated accounts contained average level of data verification, 38% of all investigated accounts contained low level of data verification). These re- sults are presented graphically in Fig. 3. 7 Fig. 4. Results of analysis of level of data verification of personal medical profiles The results show that 23% of the patients (total of 4708 person) provided reliable information in their accounts. 28% of members updated their credentials in the ac- counts. 4% of all personal medical accounts are blocked. TOOLS FOR DETERMINING VERIFIER OF PERSONAL THE DATA ADEQUACY OF DATA OF WEB- PERSONAL MEDICAL ANALYZER OF PERSONAL COMMUNITY USER PROFILES IN MIDMS MEDICAL PROFILES DATA ANALYSIS OF PERSONAL DATA AND INFORMATION MIDMS MANAGEMENT PROFILE BUILDING OF WEB-COMMUNITY USERS CCOMPONENT OMPONENT OFOF CCOMPONENT OMPONENT OF OF SETS SETS CCOMPONENT OMPONENT OFOF CCHECKING HECKING DDATA ATA OF OF FFORMATION ORMATION OFOF BBUILDING UILDING IINFORMATION NFORMATION PPERSONAL M ERSONAL MEDICAL EDICAL IINDICATORS NDICATORS PPROFILE ROFILE PPROFILE ROFILE CCOMPONENT CCOMPONENT CCOMPONENT OMPONENT OFOF OMPONENT OF OF OMPONENT OFOF M ISSING D MISSING DATA ATA IINFORMATION NFORMATION TTRACK RACK PPERSONAL ERSONAL DDATA ATA IIMPUTATION OF M MPUTATION OF MEDICAL EDICAL FFORMATION ORMATION VVALIDATION ALIDATION PPROFILES ROFILES WEB-USERS SDCH INDICATORS INFORMATION MARKERS MEDICAL TRACKS PROFILE DATA MISSING DATA GLOSSARY CONTENT PERSONAL DATA SUBSYSTEM OF INFORMATION CONTENT EXTRACTION INFORMATION SYSTEM OF MULTI-COMPUTER MEDICAL INFORMATION MONITORING WEB-COMMUNITY DECISION-MAKING USERS SUPPORT SYSTEM USERS Fig. 5. Software Complex of Determining the Data Adequacy of Personal Medical Profiles Medical Information Decision-Making Support System 8 Conclusion  the suggested method of medical information decision-making support sys- tem have been tested in 5 medical information systems.  the efficient in detecting non-valid accounts and accounts with incorrect or fake data.  the given methods simplifies work of medical staff on analysis of patients’ personal data and reduces check times. References 1. Fedushko S., Shakhovska N., Syerov Yu. Verifying the medical specialty from user profile of online community for health-related advices. CEUR Workshop Proceedings. Vol. 2255: Proceedings of the 1st International workshop on informatics & Data-driven medicine (IDDM 2018) Lviv, Ukraine, November 28–30, 2018. P. 301–310. 2. Babichev, S., Korobchynskyi, M., Mieshkov, S., Korchomnyi, O.: An Effectiveness Eval- uation of Information Technology of Gene Expression Profiles Processing for Gene Net- works Reconstruction. IJISA. Vol.10, No.7, 1-10 (2018). 3. Boyko, N., Pylypiv, O., Peleshchak, Y., Kryvenchuk, Y., Campos, J. Automated document analysis for quick personal health record creation. Proceedings of the 2nd International Workshop on Informatics & Data-Driven Medicine (IDDM 2019) Lviv, Ukraine, Novem- ber 11-13, 2019. Vol 2488. http://ceur-ws.org/Vol-2488/paper18.pdf 4. Fedushko S. Adequacy of Personal Medical Profiles Data in Medical Information Deci- sionMaking Support System. IREHI 2018. http://ieee-rural-elderly-health.com/2018/wp- content/uploads/2018/12/IREHI-Programm-1.pdf 5. Gnatyuk S., Kinzeryavyy V., Sapozhnik T., Sopilko I., Seilova N., Hrytsak A. (2020) Modern Method and Software Tool for Guaranteed Data Deletion in Advanced Big Data Systems. In: Hu Z., Petoukhov S., He M. (eds) Advances in Artificial Systems for Medi- cine and Education II. AIMEE2018 2018. Advances in Intelligent Systems and Compu- ting, vol 902. Springer, Cham. pp 581-590.