1Personal data is a key asset for many companies, since this is the essence in providing personalized services. Not all companies, and specifically new entrants to the markets, have the opportunity to access the data they need to run their business. In this paper, we describe a comprehensive personal data framework that allows service providers to share and exchange personal data and knowledge about users, while facilitating users to decide who can access which data and why. We analyze the challenges related to personal data collection, integration, retrieval, and identity and privacy management, and present the framework architecture that addresses them. We also include the validation of the framework in a banking scenario, where social and financial data is collected and properly combined to generate new socio-economic knowledge about users that is then used by a personal lending service.
Tailored and customized features are increasingly becoming more popular in IT services. These adjust offers and functionalities of services to the user preferences, interests and personal needs, generally going beyond functionality of the service itself and thus, improving it. In the banking sector, it is not an exception and for some time now new players have appeared to offer financial services based on personalization and recommendations.
Traditionally, banks have been early adopters of new
technology solutions, but mainly following a bank-centric
approach that users are rarely able to notice [
These innovative financial services create new opportunities,
but also potential threats in the industry. It is vital for banks to
understand the new directions and develop threats into new
opportunities and returns. In this sense, most of these new financial
services require personal data and financial information about users
in order to know them better and then, offer and improve services.
Here banks possess inherent competitive advantages, since they
have a large amount of customer data, transaction information, and
the capabilities to enable financing and secure services [
Well aware of this situation, in 2014 the Center for Open Middleware (COM), a joint technology center created by Santander Bank and Universidad Politécnica de Madrid, launched a pilot project intended to research, analyze and evaluate new potential opportunities and applications around personal data. Specifically, the project aims to establish a framework that allows the sharing and use of personal data among companies, and the creation of knowledge about users, while allowing users to manage and control their flow of personal information, defining who access which data and why.
In this paper we introduce the aforementioned framework which
has been called the Personal Data Framework (PeDF). The PeDF
includes mechanisms for gaining access to personal data from
several heterogeneous data sources, and integrating them to
facilitate their analysis and processing to produce and infer new
knowledge about users. This information can be provided to new
financial service providers that, as new players, do not have
sufficient personal data to offer their services. On the other hand,
there are currently tensions related to the use of personal data,
causing privacy and trust concerns in users. In this context, the
European public sector is attempting to regulate and evolve the
existing legislation to strengthen individual rights in relation to the
uses of their personal data and their privacy, while boosting digital
and personal data economy [
The remainder of the paper is organized as follows. First, Section 2 includes the technological background for each issue that covers the PeDF related to personal data: collection, integration, retrieval, and identity and privacy management. Then, Section 3 describes the PeDF architecture, and Section 4 includes the PeDF validation that we have conducted in the financial context. Finally, we present related work in Section 5, and conclude the paper by highlighting conclusions and future directions in Section 6. 2
The PeDF acts as an intermediate entity between service providers and individuals to allow the former to share and exchange existing personal data and new knowledge obtained from them which cannot be done unilaterally, while enabling users to retrieve a global view of their personal information and decide who can access which data and why. To make it possible, the PeDF has to include mechanisms for gaining access to personal data that are scattered across different service providers (data sources). When the data sources supply personal data to the PeDF, it has to be able to integrate them. This integration must allow the PeDF to provide personal data and knowledge obtained from these data to service providers (referred to as data consumers). All of the above has to be controlled by the user and thus, it requires the PeDF to include identity and privacy management solutions.
In summary, the PeDF covers four main technological issues: personal data collection, integration, retrieval, and identity management and privacy. Next, we will present the background associated with each issue, detailing its technological solutions. 2.1
Data sources can be classified into two main categories in relation to personal data access: public or private, but one source can be categorized as both, depending on the personal data concerned.
The public data sources contain personal data that are accessible in an equitable way for any entity in the public network. On the other hand, in the private data sources, the personal data can only be accessed by authorized entities. We can think of numerous examples of personal data sources, such as social networks, instant messaging services, mobile applications, and many other service providers specialized in a specific user domain such as education, banking, or e-commerce. As an illustrative example, a social network can act as a public or private data source depending on the user configuration.
There are different technologies that allow third parties to collect the personal data from data sources. For the public ones, the so-called Internet bots, spiders, or web crawlers are the most representative. These are software solutions that automatically search, access and retrieve public information on the Internet.
As regards private data sources, there are several mechanisms based on user consent that allow third parties to access the protected personal data. One of the easiest ways is the method based on data files. This kind of files contains personal data created by a user in a specific data source and can be exported by users. For example, Google allows its users to access their personal data, downloading different files2. The main problem associated with this solution is that it requires extra work for the users, since they have to be actively involved to download their files, carrying out manual tasks. Moreover, files can be easily manipulated to change their content, and therefore, the security mechanisms are weak. In order to solve this problem, a set of programming functions, protocols, and standards has appeared to automate the process: data sharing Application Programming Interfaces (APIs).
APIs have become the de facto mechanism for sharing and
exchanging personal data, since they allow different software
applications to communicate and interact directly [
In a nutshell, there is no unified API specification, each API
contains its own description, which can be poorly documented, and
therefore, understanding each one is challenging. There are some
initiatives to solve the associated API problems, such as the
OpenSocial standards [
Data integration is an old field of research that aims at combining
data from different sources and providing them in a unified view
[
Furthermore, it facilitates data analysis and processing. •
Decentralized way. Here, there is a central directory or registry and a distributed data storage. It entails little or no storage since personal data is maintained and stored by each external data source. However, personal data access is more complex and generally less efficient than the previous way because recovering data is carried out on the fly and there can be source access limitations.
The two mechanisms are complementary since the central repository of the first way can be considered as an extra storage point for the decentralized solution. Furthermore, both solutions face the challenges of corresponding personal data at different data sources, and giving them a common definition. The former entails the development of algorithms and mapping techniques that (semi)automate the correspondence process to eliminate manual tasks. On the other hand, the common definition of personal data involves establishing a standard to represent the personal data.
There is no standard or a generally adopted representation for personal data, neither the structure (format of the representation), nor even the semantic (meaning of the content). We can find many proposals for standards and proprietary solutions to define each personal data category, almost as many as there are service providers. One of the most promising solutions for integrating all these discrepancies is the use of ontologies.
An ontology is an engineering artifact made up of a vocabulary
that describes a certain reality, and a set of explicit assumptions
regarding the intended meaning of the vocabulary terms [
Over time, many ontologies have been proposed for diverse
domains including healthcare, molecular biology, or web
searching. There are general ontologies describing concepts (e.g.,
object, process and event) that are the same across different
domains, such as the Suggested Upper Merged Ontology (SUMO)
[
Finally, there are different methodologies and languages for
defining your own ontologies, such as those described in [
Personal data can be offered to third entities, and even more
interestingly, these data can be analyzed and processed to obtain
knowledge that cannot be achieved unilaterally by service
providers. The process for producing this knowledge is referred to
as user modelling in the literature [
Traditionally, user modelling is a one-sided process in which service providers autonomously collect personal data and then generate user models that satisfy their business needs in a specific domain. A user model is understood as the interpretation of a person in a specific context for an organization. It includes what the organization thinks the user is, prefers, wants, or is going to do, and comprises mainly derived and inferred data. The user model can be used to recommend new contents or services, personalize user interaction, or predict user behavior, among others.
There are different techniques to create user models, choosing one or another depends on what information is been stored and the final application of the model. Next, we point out some of the approaches that can be taken. 2.3.1
Here, a user is represented by a set of feature-value pairs. The features can be items or concepts of a domain, such as products of a shop, or links on a web site. Each of them has associated a value (usually, a boolean or real number) that indicates the attitude of a user to this feature. For example, the value can indicate whether a user has searched for a product or the number of visits to a link.
There are other approaches similar to this one such as
keywordbased, bag of words, or user-items rating matrix [
This approach is one of the simplest since its implementation
and retrieval is quite easy. It has been used by nearly every
information retrieval system [
Stereotype modelling [
The main goals of this modelling approach are to define the
stereotypes of a system and to implement the trigger techniques
that provide mapping from a specific user to one stereotype. These
include different clustering analyses, machine-learning techniques
and reasoning among others [
Classifier systems [
Semantic technologies have appeared as a way to solve
communication problems, and interoperability issues among
systems, and to provide and facilitate reusability, reliability, and a
common specification [
Using ontologies solves the polysemy problem and facilitates to
retrieve and share user models between entities. There are different
languages and techniques that allow the extraction of data from
ontologies. For example, the SPARQL Protocol and Resource
Description Framework (RDF) Query Language (SPARQL) and
the accompanying protocols [
2.4
Identity management commonly refers to the processes involved in
the management and selective disclosure of personal data, either
within an institution or between several entities, while preserving
and enforcing both privacy and security requirements. There are
different approaches to implementing identity management,
mainly: network-centric and user-centric approaches [
Network-centric approaches are based on agreements between
service providers that establish trust relationships. Each service
provider maintains its own personal data but users can link
(federate) isolated accounts that they own across different
providers to be recognized within the federated domain.
Technological standards for identity federation include the OASIS
Security Assertion Markup Language (SAML) [
On the other hand, user-centric approaches highlight user
empowerment in the governing of their personal information.
Generally, there is a third entity that is in charge of providing user
identity to service providers and the user is in the center of the
transactions, managing the sharing of personal data. Examples of
this approach are [
Identity management technologies also contribute to privacy
management by allowing users to decide on the sharing process.
However, this is not enough, as any system managing personal
information must abide by the privacy and data protection legal
framework in place, and thus fulfill a set of requirements derived
from the legal principles. For example, in Europe the main
principles include lawfulness collection and processing; gathering
specific, informed and explicit consent from data subjects; purpose
binding; necessity and data minimization; transparency and
openness; rights of the individual; and, security safeguards [
The state of the art includes a plethora of technological
solutions, each addressing a specific privacy concern, and globally
referred to as Privacy Enhancing Technologies (PETs) [
All the aforementioned technologies facilitate the access and management of personal data. However, user-centric solutions allow users to control and manage their personal data directly, bringing a better user-experience.
As described in the previous section, there are many solutions and specific technologies to handle the design and implementation of the PeDF. We have proposed a comprehensive architecture for the PeDF that considers different approaches for personal data collection, integration, retrieval, and identity and privacy management, regardless of the specific technologies and implementations. Figure 1 represents this PeDF architecture where we can distinguish its modules, and its relationships with different external data sources, data consumers, and the user.
Firstly, we have considered that there are diverse existing data sources (private or public), and crawlers on the Internet that can be linked with the PeDF to gain access to user personal data. This data source-user association can be carried out by the user through the User Manager module, or by data consumers via the Registrar module but the latter requires user consent.
Once the data sources are linked, the Collector module is in charge of obtaining personal data from them and these data have to be integrated. We have proposed two complementary approaches to carry out this integration. One is based on collecting and storing personal data, which requires a User Data Store module. The other method is based on indexing personal data, which entails a Registry module that identifies which personal data can be accessed and where they are stored.
Moreover, we have provided the PeDF with the ability to supply personal data and user models to data consumers through a Retriever module. The creation of user models entails the incorporation of different components that extract knowledge from personal data. These components have been grouped together in a main component namely Generator.
Summarizing, the PeDF incorporates seven modules: 1. User Manager. It is a vertical module that allows users to interact with PeDF to sign in, activate the incorporation of new data sources, and check and manage authorizations for access to their personal data and user accounts. It implements an identity management infrastructure and privacy solutions. 2. Registrar. This module allows data consumers to ask for the incorporation of new data sources in order to include new personal data in the PeDF. It interacts with the User Manager module to obtain the user consent.
Collector. This module is in charge of obtaining personal data from external data sources, checking user authorization. It can also include crawlers’ components that get personal data from public data sources.
Registry. It allows the PeDF to store pointers to external personal data that the PeDF is able to recover from data sources.
Generator. It comprises a set of components that allow PeDF to obtain user models from personal data. These implement different techniques of user modelling to uncover user needs, preferences, interests, etc.
User Data Store. It is a central repository that stores the personal data that is obtained from external data sources or by the Generator module. It contains different interfaces that allow the updating and refreshing of personal data.
Retriever. This module is in charge of communicating with data consumers who are interested in obtaining personal data and user models of a specific user. It interacts with the User Manager module to check user consent and with the Registry or User Data Store to retrieve the personal data requested. 4.1
We have considered two private data sources for PeDF validation: PosdataP2P service, and the social network Facebook.
PosdataP2P service [
The USC is a smart card issued by over 300 universities in collaboration with Santander Bank. It is used by 7.8 million people worldwide to access university services, such as libraries, control access (for example, to computers, campus, sports pavilions, etc.), electronic signature, discounts at retailers, etc. It can be also used to gain access to Santander Bank financial services, working as a credit/debit card linked to the holder’s saving account.
To use PosdataP2P service, USC holders have to activate the service first, providing their USC information. Then, they choose the social channels that they want to use to carry out financial transactions. Having done that, students can start making financial transactions by simply posting messages to their friends within their enabled social channels (Figure 3).
We have validated the PeDF in a banking scenario which considers a person-to-person payment service namely PosdataP2P, and the social network Facebook as data sources. Moreover, it includes a financial service called FriendLoans that uses user models from the PeDF to offer its users recommendations about microloans. It is an integration effort to provide user models that fulfill individual business needs of third entities. We have focused our work on a centralized integration based on semantic technologies, which improve the user modelling process. Moreover, we have validated the PeDF with five beta testers from our research group.
Figure 2 represents our validation to the PeDF. Here, we can observe the two private data sources (PosdataP2P and Facebook), the data consumer (MicroLoans), the user and the main PeDF modules that we have validated: User Manager, Collector, User Data Store, Generators, and Retriever. The PeDF validation applies a centralized approach where personal data obtained from external data sources are stored in a central repository. Specifically, it is based on a semantic modelling and storing, and an ontology, namely the Personal Socio-Economic Network (PSEN).
The PSEN represents the exchange of money between people and user social data. We have considered the reusing of existing ontologies, which is a must to allow semantic and syntactic interoperability. Thus, we have identified the FOAF ontology as the best alternative for representing people in a social network context and the SUMO’s financial ontology (using the OWL version) for representing the financial concepts. We have also extended them and linked the different socio-economic concepts. The nomenclature that we have used to represent the PSEN concepts is based on SUMO terms so it can be easily related to the upper ontology.
Briefly, the PSEN includes the main terms to describe people, the relationships between them, and the financial data and activities carried out between them (Figure 4). We represent people as the Person class from FOAF and we use the corresponding FOAF properties to describe their user’s demographic information: firstName, lastName, gender, age, birthday, and mbox (omitted in Figure 4 for the sake of simplicity). We also made use of the Online Account class from FOAF that allows the modelling of different web identities or online accounts of a person. We have extended it to include online payment and banking accounts. The former is devoted to service providers that allow users to carry out payment operations through the Internet, such as PosdataP2P service. It has associated a BankCard or a Financial Account class from the SUMO financial ontology that denotes where the payment will become effective. These classes have a relationship (namely, cardAccount) since a BankCard is always associated with a
class represents online banking services including financial institutions, such as Santander Bank.
To model user economic activities, we have defined a SocialInteraction class within the PSEN ontology. It includes three main properties: timestamp, channel and patient. The timestamp and channel properties indicate when and where the social interaction happens respectively, and patient designates an Entity that participates in the social interaction, i.e. the money exchange. The SocialInteraction class also has two subclasses: Transaction and Communication that have Payment and Request subclasses correspondingly. These are related to a hasPayment link that indicates whether a request for money has been paid.
In Figure 4, the rounded rectangles characterize the main concepts and the edges indicate the relationships between two classes. We have distinguished the terms of the different ontologies with darker rectangles indicated in the legend of the figure. 4.3
We have validated the retrieval of user knowledge through the
FriendLoans service, which is based on friendsourcing [
Generating a list of friends for a user requires user models that are unknown to FriendLoans, but can be retrieved from the PeDF. The PeDF has incorporated two mechanisms that allow data consumers to ask for user financial relationships and other banking information, all with the consent of the user. Specifically, the PeDF abstracts a set of SPARQL sentences and calls the reasoners which obtain and derive additional knowledge from the PSEN.
The SPARQL sentences obtain personal data and user models directly from the PSEN which can be used by FriendLoans. This information does not derive facts or inferences under the PSEN data, just data contained in it. For example, the list of friends for a specific user, if a person has carried out payments or requests for money, if a person has received money, if a person has requests for money and no associated payments, etc.
As regards the reasoners, they include the mechanisms that allow the extration of derived data. For this, we have implemented four custom rules that detect: 1) whether a user knows another user A; 2) whether a user owes money to a user A; 3) whether a user has received a payment greater than X euros; and 4) whether a user has requests for money with greater amount of money than Y euros. In the rules, the user A and the amount of money X and Y can be indicated by FriendLoans to give recommendations to its users. In this way, for the example shown in Figure 5, A will be the authenticated user Maria who needs money from her friends, X and Y could be at least 200€ or the amount wanted by FriendLoans. The results obtained from executing these rules are a set of users that fulfill all conditions. This set is not ordered since the order of execution of the rules is not predictable in the reasoner. However, the PeDF has implemented an algorithm that orders the results including tags that indicate the prioritization.
The next program listing shows an example of a rule that tags the results as the most important ones (it is indicated by the tag isFirstFor) for the user Maria (specified by the second line of the rule). The conditions of the rule are: 1) a user who has debts with Maria (defined in a function called hasDebtWith), and 2) a user has not requested an amount of money greater than 5€ with other people (defined in a function called possibleProblem). [isFirst: (?Maria psen:isTarget “true”^^xs:Boolean) (?person psen:hasDebtWith psen:Maria) noValue(?ecAct psen:possibleProblem “true”^^xs:Boolean) -> (?person psen:isFirstFor ?Maria)] 4.4
We have based our identity management infrastructure on OAuth 2.0, as it has become the de facto standard to gain access to personal data on the Web. The User Manager includes the component that manages the interaction with external sites.
Users can currently link their accounts on the PosdataP2P service and Facebook to the PeDF. The process works as follows: when a user activates a data source (i.e. Facebook), he is then redirected to the service provider site to grant the PeDF the required level of authorization. If successful, the data source delivers a token that allows access to the user profile.
As regards privacy, the PeDF has been designed to observe European privacy and data protection principles following a privacy-by-design approach. The User Manager is also the key component here, since it provides users with an identity and privacy dashboard allowing them to 1) grant/revoke consent to the collection, processing and disclosure of their personal data, 2) check the PeDF privacy policies, 3) manage the personal data known and stored by the PeDF, their sources, and the details on the disclosures to third parties as well as exercising their right to access, rectify, erase or block personal data. At the same time, the User Data Store implements security safeguards to avoid and mitigate privacy threats derived from malicious attackers or unwitting users. Finally, as regards the data minimization principle, the use of reasoners allows third parties to be limited and allows justified users to be able to query and retrieve that specified and agreed to by the data subject. 5
The PeDF is an ambitious solution that covers four main
technological challenges related to personal data: collection,
integration, retrieval, and identity and privacy management. These
have been widely analyzed separately over time in different
contexts, and we can find many researchers addressing each of
them in depth. For example, the previously cited literature [
On the other hand, there are also specialized systems, namely
Generic User Modelling Systems [
To summarize, the aforementioned solutions are rather diverse from one another, and each of them focuses on a main objective (i.e., personal data collection, identity management, and data storage). Our work is an integration effort to provide an end-to-end solution that aims at incorporating the best solutions for each issue. Our first approach is based on integrating social and financial data. To the best of our knowledge, this is the first effort in this context. 6
In this paper we have presented a comprehensive framework intermediating between users and organizations to support the seamless integration of personal data from several, distributed sources and generating advanced knowledge on users, to be shared with interested third parties, all supervised by the users who control and manage the flow of their personal data. The framework includes components for personal data collection, integration, and retrieval, as well as users’ identity and privacy management. 10 http://personal-clouds.org
The framework has been validated in a financial context, integrating social information from Facebook and a person-toperson payment service, to generate knowledge useful for a personal lending application.
Our future work includes advancing on the design of the privacy-preserving elements required to minimize the personal information retrieved by the data consumers while keeping it useful enough to fit their business needs. These developments will comprise advanced privacy enhancing technologies for attributebased credentials and database privacy.
This work is part of the Center for Open Middleware (COM), a joint technology center created by Universidad Politécnica de Madrid, Banco Santander and its technological divisions ISBAN and PRODUBAN.