Review Article A Survey On Personal Data Cloud

1y ago
5 Views
2 Downloads
2.50 MB
14 Pages
Last View : 14d ago
Last Download : 2m ago
Upload by : Luis Wallis
Transcription

Hindawi Publishing Corporation e Scientific World JournalVolume 2014, Article ID 969150, 13 pageshttp://dx.doi.org/10.1155/2014/969150Review ArticleA Survey on Personal Data CloudJiaqiu Wang and Zhongjie WangSchool of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001, ChinaCorrespondence should be addressed to Zhongjie Wang; rainy.wang@gmail.comReceived 11 April 2014; Revised 6 June 2014; Accepted 15 July 2014; Published 5 August 2014Academic Editor: Gian Luca MarcialisCopyright 2014 J. Wang and Z. Wang. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Personal data represent the e-history of a person and are of great significance to the person, but they are essentially produced andgoverned by various distributed services and there lacks a global and centralized view. In recent years, researchers pay attention toPersonal Data Cloud (PDC) which aggregates the heterogeneous personal data scattered in different clouds into one cloud, so thata person could effectively store, acquire, and share their data. This paper makes a short survey on PDC research by summarizingrelated papers published in recent years. The concept, classification, and significance of personal data are elaborately introducedand then the semantics correlation and semantics representation of personal data are discussed. A multilayer reference architectureof PDC, including its core components and a real-world operational scenario showing how the reference architecture works, isintroduced in detail. Existing commercial PDC products/prototypes are listed and compared from several perspectives. Five openissues to improve the shortcomings of current PDC research are put forward.1. IntroductionWith the flourish of cloud computing, especially of themobile computing technologies, available services on theInternet are drastically increasing and promote people’s dailylife into a “service-centric” style. In the process of servicedelivery, a great variety of heterogeneous personal data areproduced continuously. This phenomenon is in accordancewith the growing trend of “big data” in recent years. Initially,data is generated mostly by business information systemsin massive organizations and enterprises; along with theflourish of web 2.0, more and more “User Generated Content(UGC)” emerges, and today, a good deal of sensor-baseddata are automatically collected and aggregated with thehelp of the Internet of Things (IoT). To sum up, the datageneration styles have gone through three phases, that is,passive, voluntary, and automatic [1]. Especially for the lasttwo phases, personal-centric data has become the principalpart of “big data”: massive users produce personal data byvarious Social Network Services (SNS), mobile terminals, andsensors [2].In recent years, researchers pay attention to the issue ofpersonal data management, in which the effective personaldata management across services is a top priority. Efraimidiset al. [3] defined personal data as “the data created by theuser or any data about individual,” including (1) user owndata created by himself, such as social networking profile;(2) monitoring data, such as location data collected by GPSsensors in his mobile phone; (3) inferred data deduced fromthe own data and monitoring data, for example, a person’scredit score from his transaction records. Kolter et al. [4]listed various types of personal data scattered in a varietyof distributed clouds, for example, e-mail and photos storedin a web server and SMS service data stored in the mobilephone. Many other literatures have defined the personaldata; however, existing works gave the definition usually byexhaustively listing all types of personal data but there lacksa conceptual one, thereby not being able to cover all types ofpersonal data, especially in today’s situation that new servicesemerge faster and the personal data produced by the newservices is more diversified correspondingly.We consider personal data as any data that is related to aperson, not only the data produced by the person himself, butalso a (software) service or a device produced data in whichthe identification of a person is contained. Examples of theformer are documents a person drafts, videos and pictureshe took, and so forth. The latter covers more broad scope,for example, a professor’s salary information produced by his

2The Scientific World U1 dataU2 dataU3 data···U1 dataU2 dataU3 data···U1 dataU2 dataU3 DC (U1 )Figure 2: PDC: User-centric personal data management.Figure 1: Service-centric personal data management.university’s human resource service, a patient’s health recordproduced by his wearable devices and medical devices of ahospital, and a traveler’s air travel records produced by theonline booking services of different airlines. If we considerthe person himself as a “human service,” the former type ofpersonal data may be regarded as a special one of the latter.Here we give a uniform definition:Definition 1. Personal Data is the data produced by anyservices (not only software and web-based services, but alsohuman and device-based services) around a person duringthe interactions between the person and the services, oncondition that if the person’s identification is removed fromthe data, the data will become meaningless.Due to the heterogeneity and distributedness of personaldata, personal data management exhibits very different characteristics compared with enterprise data or service datamanagement. For the moment, most of the personal data isgoverned by the provider of the service that produces thedata and the data is stored in the cloud of the provider. Ifa person uses twenty services in his daily life, his personaldata is consequentially distributed among twenty logicallyindependent clouds. The person who essentially owns thedata has limited privileges only in each service domain butno full authorities to share his personal data across theboundaries between different services. In this case, the personcould never get a unified global view on his own personaldata. This deprives the right of users as the owner of their data.Especially, in most cases, a person’s daily requirements willspan across multiple services, and the isolation of his persondata hinders the possibilities of autonomic collaborationsamong services around him. Figure 1 shows such servicecentric personal data scenario.To address this challenge, some researchers proposed anidea called “Personal Data Cloud (PDC)” to collect and storethe personal data of a user in a centralized cloud. PDC, a SaaSapplication deployed on a specific PaaS platform, plays therole of personal data management in a holistic way. Ideally,services around a person will send the generated personaldata directly to his PDC; if not, a dynamic personal datacollection component is required to facilitate the synchronization between the cloud of services and PDC of the user.We call it a “user-centric” model shown in Figure 2.In Figures 1 and 2, Appi refers to an application deployedon the mobile terminals or accessed via web browsers, 𝑆𝑖 isa software service deployed on a cloud or a physical servicedelivered by a device, and the cylinders are the data storagesin the clouds. In the service-centric scenario, each serviceis connected to its own cloud and the personal data ofmultiple users of this service is stored in the same cloud.Comparatively, in the user-centric scenario, the cloud belongsto the user himself and all his personal data (no matterwhich service produces the data) are stored in his PDC. Theadvantage is evident: the isolations among different servicesare broken and the inherent relations between the user’spersonal data are recovered and governed in the PDC. Underthe support of PDC, the collaborations between servicesaround a person become possible.This viewpoint is widely endorsed by literatures. Munet al. [5] thought that “user-centric” is the gravity shift ofinformation management from organizations to individuals.Technically, this scenario stores personal data in one centralcore, with domain-specific services plugged into the core, andit is the user that owns the full authority of controlling hisdata. Ardissono et al. [6] put forward the concept of personalcloud, an infrastructure providing an abstraction level overvarious individual applications and services. Being a unifieddata management environment, the personal cloud offerscomplementary functions instead of just linking separateapplications and workspaces. Kirkham et al. [7] also proposeda similar idea and believed that such centralized personal data

The Scientific World Journalcloud would effectively enable service collaboration aroundusers.As this is an emerging issue in both research andpractice, this paper makes a brief survey on PDC by making an elaborate analysis and summary of related literatures. The objective is to give service and cloud computing researchers/practitioners a global view about the latestresearch and development on PDC. Remainder of this paperis organized as follows. Section 2 gives the classification andrepresentation of personal data. A reference architecture ofPDC (including its primary components) is introduced inSection 3. Section 4 lists some existing commercial PDCproducts/prototypes and compares them from seven perspectives. In Section 5, some open issues about PDC are discussed.Section 6 concludes the paper.2. Background of Personal Data2.1. Value and Significance of Personal Data. The value andsignificance of personal data have been fully recognized.(1) Personal data is a partial representation of personalInternet footprint which gradually grows along witha person’s daily usage of various services and mobiledevices over a period of time. Website accessingrecords, keywords entered in a search engine, browsing history in an e-Business website, and so on allbelong to the personal footprints [8]. These typesof data are unintentionally generated by users butare carefully tracked and recorded by search enginesand service providers. Analytical tools for recording, aggregating and analyzing the footprints fordeep understanding on user behaviors, for example,NM Incite, Social Mention, SocMetrics, Traackr, andTweepi [9], have been widely adopted.(2) Personal data is a partial representation of personale-history. In the electronic age, people’s daily life isfull of intensive interactions with various services,and the generated personal data constitutes his ehistory which is always growing. How many citieshave I visited in my past life? What kinds of bookshave I bought from multiple online bookstores? Howmuch investment income have I attained from fourbanks in the last five years? There are many suchquestions, but they are all difficult to be answeredquickly and accurately. Having effective personal datamanagement, people possess the ability of reviewingand summarizing their own history from variousperspectives [10].(3) Personal data is a partial representation of personalhabits and preferences. Similar to the footprint, personal habits and preferences are prolifically embedded in personal data, too [11], for example, my favoritebooks/music, my wish list, my comments on sometopics, my blogs, and so forth [12]. Without PDC,each service provider is responsible for collectingsuch data and analyzing user habits and preferencesto make accurate service recommendations but is3limited to only one service domain. If PDC was used,such analysis could cover the full-scale personal datafrom various services and will be more effective andprecise.(4) Personal data enables completely personalized service collaboration. Traditional service collaborationis usually dominated by service brokers in a publicservice platform (e.g., eBay’s e-commerce platformand Expedia’s online travel platform), but this wayusually offers standard collaboration patterns withthe limited degree of personalization. If PDC exists,the personal data represents a user’s history andpreferences (namely, his personalized requirementson services); therefore, it is easy to conduct thecompletely personalized collaborations between theseservices, and the service brokers are no longer necessary [7, 13]. Some commercialized services suchas ifttt.com and Google Now also support the personalized service collaborations in a user-centric wayinstead of traditional broker-centric one.2.2. Features of Personal Data. The following four distinctcharacteristics jointly differentiate the personal data fromother types of data.(1) High degree of dispersion: referring to the fact that thepersonal data is scattered in a wide range of IT environments (clouds, mobile devices, etc.) throughoutthe hardware and software and a variety of serviceproviders [13, 14], thereby it is difficult for data ownersto uniformly manage their personal data.(2) High degree of heterogeneity: referring to the fact thatpersonal data is composed of a variety of morphologies, with different data types and granularities, andabove all, with different semantics representations.The corresponding challenge is the syntax and semantics unification [15].(3) High degree of correlation: referring to the fact thatthere are close correlations between different parts ofpersonal data which are originally stored independently and isolated with each other. This is becausethese data describe the person’s life from differentaspects, and such correlations do not depend onwhere they are stored and who they are managed.To recover such intrinsic correlations, ontology andLinked Data are frequently adopted to correlate thepersonal data released on the web by using URI andRDF [16].(4) High degree of privacy: referring to the fact thatpersonal data should be shared in a strict and limitedscope with other people/services. Personal data isvulnerable to be attacked, and excessive openness willresult in a lot of privacy and security problems. Muchresearch work such as [7, 17] focuses on the privacyof personal data to ensure that the sensitive personaldata processing takes place within the user’s PDCinstead of a third-party server.

4The Scientific World JournalTable 1: Multidimensional classification of personal data.DimensionCategoriesExamplesFormat(1) Document(2) Multimedia(3) Web page and fragment(4) EmailDOC, PPT, spreadsheets, and so forthImages, videos, audio, and so forthSearch keywords, visited links, cookies, and so forthGmail, Yahoo! Mail, and so forthBusiness data stored in domain-specific services, suchas orders, calendars, wish lists, and so forth(5) DatabaseSource(1) Personal devices(2) Services(3) Social network(4) Sensors(5) The person himselfReferencesPC, smart phones, mobile devices, tablet, and so forthWeb applications, and so forthFacebook, Twitter, blogs, and so forthGPS, thermometer, wearable devices, and so forthEmail, work schedule, documents, pictures, video,audio(1) Metadata(2) Instance dataThe descriptions of personal dataThe contents (instances) of the metadataPreferences on books, music, cities, friends, wish list,(1) Preference dataand so forth(2) Communication recordSMS text, phone records, address book, and so forthVisited websites, search keywords, social comment logs(3) Web footprintsand social graph, and so forthHeight, weight, published papers, education/careerSemantics and functions (4) Personal profileexperiences, exam performance, and so forthBank account and transaction records, flight and hotel(5) Consumption service record orders, car rental orders, supermarket records,e-commerce transaction record, and so forthPersonal salary records, household energy record,(6) Public service recordpersonal credit, and so forth(1) Local/desktop storageFiles located on personal computers and devicesStorage location(2) Distributed cloud storageData stored in the cloud of a service(3) Centralized cloud storageMany personal data centralized stored in a public cloudAbstraction level2.3. The Classification of Personal Data. Researches havemade elaborate classifications on personal data in terms ofdifferent criteria. This section summarizes previous work andgives a comprehensive classification. It is shown in Table 1.The first dimension is the format which the personaldata externally exhibits in, including documents, multimedia,web pages/fragments, email, and database. The second oneis the source where the personal data is generated, includingpersonal devices, web-based services, social networks, sensors, and the person himself. The third one is the abstractionlevel of personal data, including meta- and instance data.The fourth one is the semantics and functions and is themost complicated one, including the preference data, webfootprints, and consumption and public service record. Thelast one is from the location where personal data is stored,including local/desktop data, distributed cloud based data,and centralized cloud-based data. Examples for each dimension are shown in the third column, with related literatures inthe last column.2.4. Views of Personal Data. Because of the complexity andthe high volume of personal data, it is difficult to visualize thedata all at once. Here we give five views to help decompose the[3–5, 9, 10, 12, 13, 18][5–8, 14, 19–22][11, 14–17][19, 23–32][20, 33–36]whole personal data into small parts so as to achieve clearervisualization effects and better understanding on the data. Itis called “data projection” being adopted in data visualizationdomain and so does in the personal data research. The fiveviews are listed as follows.(1) Time (when): it organizes those personal data havinga timestamp attribute in the form of the timeline. Theunit of time might be a day, a week, a month, ora year, depending on the time granularity that theuser is concerned about. Each data item is annotatedonto the timeline in terms of the timestamp it owns,and the timelines will show different time granularity,for example, year, month, week, day, and so on.Data without any timestamp is not visualized. Forexample, a timeline is used to show the personalenergy consumption of both household and businessactivities by the time view [37]. Further, personal datais classified into three tenses: past, present, and future.(2) Location (where): it organizes the personal data having a location attribute in the form of a geographical map. Many personal data have location-relatedattributes, so it is convenient to visualize the data on

The Scientific World Journala map with latitude and longitude coordinates. Forexample, a world map is used to present the travellinglocation and route of users [37].(3) People (with whom): it organizes the personal datahaving some socialization attributes that direct toother persons. In other words, these data representthe user’s social networking with others. Usually adirected graph is adopted to show the data projectedin this view [4, 38].(4) Belonging (what): it views the personal data standingfor a virtual or physical belonging of the user, forexample, air miles, books, cars, and clothes. It isusually visualized in the form of a list.(5) Finance (how much): it views the data having someattributes with economics significance, that is, the datapointing to a specific financial transaction [10]. Forexample, a transaction record from PayPal, a purchaseorder from Amazon, and a credit card bill fromCitibank. This view is usually visualized in the formof income and expenses curves.It is important to note that each personal data item mightfall into multiple views. For the purpose of personal datavisualization, it is necessary to design for each view, and thecombination of two different views, and so forth. Figure 3shows some examples of the personal data visualization,where Figure 3(a) is the time view, Figure 3(b) is the locationview, Figure 3(c) is the people view, and Figure 3(d) is thefinance view.2.5. The Semantics Correlation between Personal Data.Although the personal data are aggregated from multipleservices, they are inherently correlated by the user. This iscalled data correlations. For example, an activity “A businesstrip to Alaska for attending 2014 CLOUD conference” inGoogle Calendar is directly related to a flight order in Expediaand then related to a transaction record in PayPal, and soforth.Data correlation will bring many benefits to the users.If we correlate personal data from various sources andlink a wide variety of personal data in the web, the queryefficiency could be speeded up [39]. Data correlations couldbe expressed in the form of static explicit declarations or in arelational data base system [40, 41].But due to the high degree of dispersion of personal data,most of such correlations have disappeared. The recoveryof semantics correlations after person data is collected is achallenging issue. Actually this is also the ideal of SemanticWeb community and some feasible techniques such as Linkeddata have been put into practice in recent years.2.6. Semantic Representation of Personal Data. Ontology andLinked data are the popular approaches for the semanticrepresentation of personal data. Ontology defines a set ofdomain-specific concepts, attributes, and relations using ashared vocabulary [42]. An example is from [43] where anovel method is proposed to describe the metadata and5instances of personal data in the form of ontology andprovided an intelligent way to manipulate the data.Linked data is an effective technique to interlink, share,and publicize various web resources by predefined ontology,built upon standard Web technologies such as HTTP, RDF,and URIs; thereby they can be automatically manipulated bycomputers. This enables personal data from different sourcesto be connected and queried efficiently, too [44].3. Personal Data Cloud (PDC)Based on the survey on personal data, we summarize theresearch progress on PDC and present a reference architecture of PDC.3.1. Synonyms of PDC. PDC is a term proposed in this paperwith the implications of collecting, aggregating, storing,indexing, correlating, and using the personal data. In thedomain of personal data management, researchers focus onthe same objectives but have employed different terms, suchas the following.(i) Personal Information Management (PIM) focuseson the acquisition or creation, store, organization,maintenance, retrieval, usage, and distribution of thepersonal information [45, 46].(ii) Personal Data Spaces (PDS) is an abstract datamanagement technique aiming at personal data integration, based on existing matching and mappinggeneration techniques [17, 47].(iii) Personal Data Store (PDS), or called personal datavault or locker, is a service allowing an individualstore, manage, and deploy their key personal data ina highly secure and structured way [5, 7, 48].(iv) Consumer-Centric Cloud Portal (C3P) is a middleware acting as an intermediary between Apps andservices and assists Apps access the personal data incloud in a device-, time-, and location-independentway [49].(v) Personal Cloud Butler (PCB) is a service that providesa safe haven for personal digital assets and supportssharing with fine-grain access control [50].(vi) Personal Cloud (PC) is a similar service allowingusers access their personal data across multipledevices [51].The reason why we use Personal Data Cloud (PDC) tounify this miscellaneous terms listed above is straightforward:firstly, the managed object is “Personal Data;” secondly, themanagement of personal data is more inclined to a centralized cloud environment; thirdly, the management issuesshould cover the full lifecycle of personal data.3.2. Reference Architecture of PDC. Essentially, a PDC is aSaaS application deployed on a cloud. A reference architecture is necessary for PDC developers to plan its maincomponents and their interconnections.

6The Scientific World Journal2011Pay house energy2012Purchase an ipad2012Travel to America2013Visit websitesYearlyJan 2011 Feb 2012Open bank accountJan 2014Jul 2012 Jan 2013Oct 2012Send SMS textShare pictures in twitterJan 2012MonthlyJanFeb Mar Apr May Jun JulJun 3, 2012Jun 6, 2012Send an emailPublish papers12Sep Oct NovJun 11, 2012Take examAugJanDecDaily3457689101112141315(a) Time viewAttend EnglishclassHave a dinner withfriendsPurchase newfurnitureTake the subwayHave an appointmentwith friends(b) Location viewGenderEmailBirthday rEmailDavid BirthdayAddress lary bills4000GenderEmailSmith BirthdayAddressJack(c) People view123456789101112MonthExpensesIncome(d) Finance viewFigure 3: Personal data visualization for four different views.As shown in Figure 4, PDC has a multilayer architecture supporting the seamless integration between the Appsinstalled on mobile terminals and a set of PDC servicesdeployed on the cloud. This architecture is proposed bythe synthesis of the personal data management frameworkspresented by the literatures mentioned in Section 3.1.Here we give a brief introduction to each layer.(1) Personal Data Ontology. It is an extensible ontologydefining a set of standard terms (classes, attributes,and relations) that covers various service domains.It offers the abstraction of various types/sources of

The Scientific World Journal7Data storage layer (linked data)PersonalFundamental service layer (data ationData nagementData statusmanagementPDC open APIProxy forterminal appsServicecoordinatorontologyPortal forpersonal , web)UserFigure 4: A reference architecture of PDC.personal data and is intended to be completely independent of the physical representation of personaldata.(2) Data Storage Layer. It is a centralized data repositorywhere the meta and instances of the personal data arecentrally stored. Either the metadata or the instancedata is annotated by the Personal Data Ontologyso that their semantics is unified and the potentialsemantics relations are recovered. All the data isindexed and represented in the form of Linked Datawhich facilitates the convenient query and navigation.(3) Fundamental Service Layer (Data Engine). It is thecore of PDC and composed of a set of fundamentalservices.(i) Service Registration component allows users toregister the services they are using to PDCso that the personal data that these servicesproduce is to be imported to PDC for the unifiedmanagement.(ii) Data Importation component enables the (semi) automatic importation of the personal dataproduced by the registered services into PDCand the data synchronization between servicesand PDC if the same data in either side wasupdated.(iii) Semantic Annotation component is to establishthe semantics mapping between the metadataimported from services with the Personal DataOntology for semantics unification.(iv) Data Correlation component is to manage thesemantics correlations between personal dataproduced by different services so that they arerepresented as Linked Data with the help ofPersonal Data Ontology.(v) Privacy Control component is used to set up theprivacy rules/policies on the personal data, forexample, what classes, attributes, and relationscould be accessed by which of the externalservices and which of the other users, therebyprotecting the data privacy. This issue is to bediscussed in Section 3.3.(vi) Data Status Management, Event Management,and Triggers are the three components enablingthe PDC-based service collaborations. DataStatus Management is responsible for monitoring the dynamic changes of personal dataand then generating the corresponding events;Event Management consolidates all the generated events in a queue; and Triggers try toidentify the potential collaborations, distributethe related events to external services or mobileapps, and then trigger the collaborations.

8The Scientific World Journal(4) PDC Coordinator. As each user has his own PDC,the PDC coordinator enables the communicationbetween multiple PDCs so that the social servicecollaborations between different users are established.(5) Service Coordinator. It is responsible for the coordination between the services that have been registeredto PDC when the Trigger component identifies thepotential collaborations between them.(6) Proxy for Terminal Apps. The potential collaborationoccurs not only between services, but also possiblybetween the apps in mobile terminals. Each terminalapp has a proxy on PDC and could be triggered by theproxy through callback mechanism. In other words, achange of personal data would lead to the executionof some actions offered by the apps.(7) Open API. It facilitates bidirectional data exchangebetween PDC and various terminal apps, allowing theapps access the data in PDC in a standard way.(8) Portal for Personal Data Visualization. It is a GUIwhere users browse and query their personal datain selected view(s) and tense(s) (discussed inSection 2.4). Data is graphically visualized.(9) Apps. This refers to the various terminal apps.It is noted that not all above components have beenimplemented by existing works. The PDC architecture is stillan open issue both in research and practice.3.3. An Operational Scenario of the PDC Reference Architecture. To illuminate how the PDC reference architectureworks, here we give an operational scenario. Suppose thereare two users named Jack and Lily who have their ownPDC, and they use a set of services including TypoWeather(a weather forecast service), EatThisMuch (an automaticdiet planner service), MyClean (a maid cleaning service),HealthLoop (a medical service to monitor and communicatewith patients during the recovery process),

In the process of service delivery, a great variety of heterogeneous personal data are . data hinders the possibilities of autonomic collaborations among services around him. Figure shows such service- . is usually dominated by service brokers in a public service platform (e.g., eBay s e-commerce platform and Expedia s online travel .

Related Documents:

Amendments to the Louisiana Constitution of 1974 Article I Article II Article III Article IV Article V Article VI Article VII Article VIII Article IX Article X Article XI Article XII Article XIII Article XIV Article I: Declaration of Rights Election Ballot # Author Bill/Act # Amendment Sec. Votes for % For Votes Against %

Article 27 Article 32 26 37 Journeyman Glazier Wages Article 32, Section A (2) 38 Jurisdiction of Work Article 32, Section L 43 Legality Article 2 3 Mechanical Equipment Article 15, Section B 16 Out-of-Area Employers Article 4, Section B 4 Out-of-Area Work Article 4, Section A 4 Overtime Article 32, Section G 41

Jefferson Starship article 83 Jethro Tull (Ian Anderson) article 78 Steve Marriott article 63, 64 Bill Nelson article 96 Iggy Pop article 81 Ramones article 74 Sparks article 79 Stranglers article 87 Steve Winwood article 61 Roy Wood art

1 ARTICLES CONTENTS Page Article 1 Competition Area. 2 Article 2 Equipment. 4 Article 3 Judo Uniform (Judogi). 6 Article 4 Hygiene. 9 Article 5 Referees and Officials. 9 Article 6 Position and Function of the Referee. 11 Article 7 Position and Function of the Judges. 12 Article 8 Gestures. 14 Article 9 Location (Valid Areas).

article 22, call time 41 article 23, standby time 42 article 24, life insurance 42 article 25, health benefits 43 article 26, work-related injuries 51 article 27, classification 55 article 28, discharge, demotion, suspension, and discipline 58 article 29, sen

Section I. Introductory provisions Chapter 1 General Provisions (Article 1 - Article 9) Chapter 2 Voting rights (Article 10 - Article 11) Chapter 3 Electoral Districts (Article 12 - Article 17) Chapter 4 The register of voters (Article 18 - Article 25) Ch

Article 9. Conditions of Operation Article 10. Disciplinary Actions Article 11. Penalties Article 12. Revenues Article 13. Local Governments Article 14. Miscellaneous Provisions Article 15. Additional Restrictions Related to Fair Elections and Corruption of Regulators Article 16. Additional Contracts: Proposition Players

central fleet management 07/01/2022 — 06/30/2025 reformatted july 2021 table ofcontents preamble 4 article 1: recognition; 5 article 2: non-discrimination 6 article 3: maintenance of membership 7 article 4: dues deduction 8 article 5; seniority 9 article 6: promotions and transfers 11 article 7: wage rates 13 article 8; hours of work and .