Keywords

1 Introduction

In a scenario where the Internet has established itself as an environment where we can make complex researches, query a huge amount of databases and search for information, the social networks have become a choice of information source. Social networks are able to generate great impact on a particular subject, regardless of the provenance of information, whether from official sources or simple user status update. This kind of system provides first-hand data, but one pressing problem is to distinguish true information from misinformation and rumors. In many cases, social media data is user generated and can be biased, inaccurate, and subjective [1].

Many users are not ready to determine the credibility of online information. Given this statement it is essential to encourage the development of people skills to make content evaluation on the Internet, however how people seek information and the types of information they seek is a complex interplay between characteristics of the information seeker, the properties of the information they are looking for, and the systems in which they seek information [2]. Considering the behavior and characteristics of the users plus the information that they face as variables, we focused on developing tools for social networks intended to support verification of content.

The aim of our work is to analyze, build mechanisms and subsequently make the operationalization of auditability characteristics as tools for information validation, using references from interaction design, in order to obtain better usability in such systems. Successful tools for community generated content need to address the lack of transparency [3]. To create such tools we applied concepts of transparency on information systems headed up by the provision of audit capacity of spreaded content. In other hand we aligned this theme with Human-Computer Interaction (HCI) research guidelines.

This article is based in an exploratory approach and is structured as follows: In the Sect. 2 we present our analysis about the non-functional requirements related to a catalog of information auditability in social networks and the HCI foundations to build tools based on these requirements. Some of the suggestions of tools, main purpose of this study are explained and presented as prototypes in the Sect. 3. The conclusion of this article, citing the next stages of work is available in Sect. 4.

2 Auditability and Human-Computer Interaction

2.1 A Catalog to Define Auditability Characteristics

To promote the information audit in social networks we need to deal with non-functional requirements (NFR). The identification and proper expression of NFRs are essential to the understanding and reasoning of the impacts of further design decisions [4]. To set these NFRs we can use a catalog [5].

The Catalog of Transparency developed by Cappelli [6] is a pillar for this research, especially as regards the characteristic of auditability and its definition. Auditability is defined as the ability to examine carefully for accuracy with the intent of verification [7]. This characteristic is identified through the assessment of practices that implement explanation, traceability, verifiability, validity and controllability of information. We use these definitions as a starting point to formulate a catalog for information auditability characteristics in social networks.

2.2 Relating HCI Definitions and Auditability Characteristics

After define auditability characteristics, the next step of this work consists in specify a set of operationalizations and implementation mechanisms that can implement these characteristics um software. To identify them we applied the foundations and practices of HCI area.

To help us with the challenge of listing non-functional requirements related to auditability, turning it on tools for evaluation of information by users, we focused our research on a literature review related to experience in the use and interaction with web interfaces, in the construction of quality systems and in recommendations for best practices defined by international standards [8]. After this step we reached, in a first approach, 15 characteristics that can be explored in order to minimize the problems of the lack of auditability of information in social networks. These characteristics are: accountability, adaptability, availability, clarity, completeness, composability, controllability, correctness, decomposability, dependability, extensibility, traceability, uniformity, validity and verifiability. The implementation of some of these characteristics will be reported in Sect. 3 of this article.

According to Baranauskas [9], the HCI research is based on the belief that we have people using interactive systems. These interactive systems should be designed following requirements, capacity and user preferences for an interaction with usability. People do not have to change to suit the system, the system must be rather designed to suit their requirements. We need to develop tools that besides providing the audit capacity of information need to be adapted to any kind of user. Some researches [10] found that users judgments were based on two basic criteria: the characteristics of the information “objects” that they found online and the characteristics of the source of online information. The characteristics of the information objects that participants were concerned about included the type of information object, its content, its presentation and its structure.

After a review of numerous models (created by many authors) of evaluation information, Metzger [11] proposed in an article a common checklist to credibility assessment. In this way the tools that follow the suggestions of the checklist would be able to help social network users acquire the skills needed to find credible online information and to evaluate existing theoretical models and user-based strategies of credibility assessment. Table 1 shows the relation of some functional requirements based on HCI guidance and the auditability characteristics.

Table 1. Relationship between HCI based items and the Catalog of Auditability
Full size table

3 Reviewing Tools Prototypes

Using Fig. 1 as a guide we can see the number of items that we will cover in this prototype. In the tooltip 1 we bring as a requirement, a feature that permits the user to assign other profiles to verify the user’s reputation and online social presence. In this example we use LinkedIn, because it is a social network for job search and is essential that the users are presented as they really are. This feature enhances the positioning information produced by the users and is related to the composability, which describes the gathering of more data from other partitions adding it to the existing information and traceability.

Fig. 1.
figure 1

A social network with several prototypes of suggested tools

Full size image

It is indispensable that all shared information is marked with a timestamp as shown in the tooltip 2. The existence of this option shows whether the information is current and contextualizes the users when they check information. For the verifiability is desirable that the information has a date and time, so we can use as a filter to legitimize that content and traces it. The user qualifications as information producer can be considered as a reference by other users. The tooltip 3 shows the importance of having a social network account with complete register and profile image. The dissemination of information on social networks is enhanced by sharing of external links. Track these links can give us tips on the information contained in these destinations and guide users about what they can find. The availability of the link is another item considered on our list (tooltip 4) because with no data available for analysis we do not have content and cannot trace related information. The tooltip 5 deals with the correctness and clarity of information. Well written, clear and concise information show the concern and commitment of the source. Provide tools such as spell checking will help users to improve the information that they are producing. The main functional requirement related to the prototypes will be the capacity to alert a user when information on social network contains any hint of misinformation based on NFR concepts that we mapped.

4 Conclusion

In this work we present some guidelines for development of interfaces with tools that support transparency and auditability in social networks. We visited the concepts in the HCI area to help us with this approach. The directions of this work focus in a way to take advantage of different forms of social networks, seeking their auditability. Our next steps are: continuing the analysis of the operationalizations and mechanisms and a tool construction implementing the prototype. Before develop the tool we will check these prototypes using Nielsen Heuristics [12]. Another suggestion for the prototypes review is submit them to the approach of five traditional evaluation criteria: authority, accuracy, objectivity, currency, and coverage. These criteria have their origins in the world of print, but are universal criteria that need to be addressed regardless of the information being evaluated [13].