47 Semantic Web End-User Tasks Roberto Garcíaa a Universitat de Lleida, Jaume II 69, E-25001 Lleida, Spain Abstract In order to make the Semantic Web reach “real world” end-users it is important to consider Semantic Web usability. User-Centered Design from the Human-Computer Interaction com- munity might help in this respect. First of all, the user must be defined, together with the context. Then it is possible to study user tasks. We focus our study of Semantic Web user tasks in end- users and Semantic Web online applications, trying to contribute to establishing some UCD guidelines that help the adoption of Semantic Web applications. However, we consider existing analysis for Web systems and even online information systems in general in order to avoid constraining our view to the current state of development of the Se-mantic Web. The proposed set of end-user Semantic Web tasks is Search, Browse, Annotate, Mashup, Map, Share, Communicate and Transact. They are used in order to study an existing Semantic Web platform and an application based on it. This allows putting the tasks into practice and relates them to some interaction patterns. Future work continues in this line, trying to connect the identified patterns with existing and new interaction pat-terns in order to contribute additional guidelines for UCD Semantic Web applications development. Keywords 1 Semantic Web, Human-Computer Interaction, User Task, User Experience, User Interface 1. Introduction process, and the objective is to get usable and accessible products. In the context of HCI, User-Centred Design The Semantic Web has been around for (UCD) proposes facing the development some time and many people are asking why it process of interactive systems focusing on the has not taken off as quickly as the World Wide user and considering the Quality in Use, Web did [1]. One of the main impediments is standardised in ISO/IEC 25010:2011 [4]. The that it is not reaching the end-users, who can proposed development process starts with a give it the required critical mass for widespread characterisation of the target users and the tasks adoption. End-users find Semantic Web they carry out with the interactive system in applications very hard to use, it is difficult even order to meet their needs, and the metrics it for researchers and practitioners working in the proposes to evaluate the quality in use have Semantic Web field [2]. been extended to Semantic Web exploration Once Semantic Web technologies seem to tools [5]. be quite mature, in order to facilitate its The tasks supported by the early Web are adoption, it is time to focus on the face now neatly defined and are becoming part of Semantic Web applications show to users [3]. Web developers’ common practice, making it Human-Computer Interaction (HCI) is a relatively easy to develop tools adapted to these multidisciplinary effort to improve the human- tasks following a UCD approach. Knowledge computer interface. The focus is placed on the about tasks in the Semantic Web is much less user, i.e. to consider user needs from the clear due to its novelty, but it is necessary in beginning and through all the development International Semantic Intelligence Conference, February 25–27, 2021, New Delhi, India EMAIL: roberto.garcia@udl.cat ORCID: 0000-0003-2207-9605 © 2021 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). CEUR Wor Pr ks hop oceedi ngs ht I tp: // ceur - SSN1613- ws .or 0073 g CEUR Workshop Proceedings (CEUR-WS.org) 48 order to be able to take a user-centred view on 2. Related Work the Semantic Web [6]. However, before tasks can be identified, the Due to the importance of a clear definition first step is to determine who the intended users of the user tasks when developing interactive are and their context. There might be the systems, especially if an UCD approach is temptation to say that users are "everyone", but followed, there are many studies of this kind. UCD recommends dividing the target audience Usually, they refer to the user tasks for a into groups. Considered that the objective is to concrete application. However, there are also facilitate Semantic Web adoption among end- studies that consider a range of applications in users, this seems the target audience to focus order to help defining guides or analysing on. On the other hand, the context is any online common practices. application based on Semantic Web When analysing existing work, we have technologies, which seem the best channel for a considered different studies at different scale widespread adoption of the Semantic Web. levels, from the more general online In the context of this paper, end-user is information systems, through Web information defined as a user with no or limited knowledge systems, to Semantic Web applications in about Semantic Web technologies and general and finally some specific Semantic methodologies. We do not include in this user Web scenarios. The objective is to not pass profile domain experts, that might be also users through potential tasks that might be considered with limited knowledge about the Semantic in the context of the Semantic Web but that Web but who have specific needs related with haven't been considered yet in that context. the development of ontologies. Therefore, starting from the broader context, Once we have characterised the kind of Heath et al. [8] propose a set of user tasks users Semantic Web user we are interested in, it is carry out on-line with information systems. time to determine the tasks. As we are They take a quite broad point of view as they considering the Semantic Web as a whole, and include the Web but also other Internet not a specific Semantic Web application, the application like electronic mail or instant tasks should be generic and broad enough to messaging. The list includes Locating, accommodate tools that are not yet Semantic Exploring, Grazing, Monitoring, Sharing, Web enabled but that might be so in the future. Notifying, Asserting, Discussing, Evaluating, In order to make the range of tasks broad Arranging and Transacting. enough and avoid constraining our view to the Locating is about users looking for current state of development of the Semantic something known or expected to exists. Web, the focus should be broader than Exploring refers to gathering information to existing Semantic Web applications and studies gain understanding or background. Grazing is of Semantic Web user tasks. We also consider moving speculatively without a specific goal. user tasks in the context of the Web and even in Monitoring is about checking known source the context of information systems. This makes expected to change. Sharing refers to making it possible to check the consistency and something available to others. Notifying is coverage of the proposal. The range of user informing others about something that happens. tasks studies under consideration is presented in Asserting is about making statements of fact or Section 2. opinion. Discussing refers to exchanging From the analysis of these existing studies, information on a topic with others. Evaluating which range from Semantic Web to information is determining if some information is true or systems user tasks, we build our proposal of a alternatives. Arranging is about coordinating set of generic Semantic Web end-user tasks, with third parties. Transacting is transferring which is presented in Section 3. Then, in order money. to study these tasks deeper, we have then put Getting into a more specific context, Kellar them into practice in Section 4. We have et al. [9] present a quite complete summary of isolated them in the context of Rhizomer [7], a studies about Web information systems user platform that makes it possible to develop tasks. In their web information task Semantic Web enabled sites with content classification, they identify a set of user tasks management capabilities. Finally, Section 5 that are classified in three information goals: presents the conclusions and the future work. information seeking (Fact Finding, Info 49 Gathering and Browsing), information Action- to build a exchange (Transacting and Communicating) oriented personalized portal and information maintenance (Maintaining). tasks to manage research Fact finding is usually a short task that tasks stands for looking for specific pieces of Information to share pictures sharing tasks with friends and information. Information Gathering involves family the collection of information and Browsing is a Content Content to add new books serendipitous task where users have no specific curators update tasks to a catalog of goal in mind. In relation with the information published books exchange goals, there are Transacting, which and edit the stands for performing an on-line action that metadata of often involves user/password authentication, previously added and Communicating, connected to web-based ones communication, e.g. e-mail or blogs. Content to provide The last goal defined by Kellar is distribution information to information maintenance that includes just one tasks museum visitors user task, Maintaining. This task is about Ontologists Ontology to reorganise a update tasks library editing web resources in order to make them categorisation work properly, e.g. no broken links, and update scheme them. Kellar et al. also consider a potential task, Ontology to map between Monitoring as returning to a previously visited creation & different medical page in order to obtain updated or dynamic mapping ontologies information. However, they do not consider a user tasks per se, but a task dimension, re- On the other hand, Mäkelä et al. [11] present occurrence, that might be a characteristic of any three very generic tasks that need to be handled of the previous user tasks, especially Browsing, in any information system with semantic Transacting, Fact Finding and Information capabilities: Gathering. • Semantic Content Consumption is about If we concentrate now on Semantic Web consuming semantic content when users applications, Battle [10] provides starting are searching, browsing or other tasks points for describing Semantic Web users and like aggregating an RSS syndication their tasks. ¡Error! No se encuentra el origen service. de la referencia. shows the three high-level • Content Indexing stands for the tasks categories of Semantic web users and the kind where ontologists or end-users produce of tasks commonly associated with each group, semantic metadata by indexing and together with an example for each task. Her publishing content with references to characterisation of end-users is “users that do shared vocabularies. End-users may play not know what the Semantic Web is and that do a role of content indexers when they are not care as long as they can get what they need sharing videos or blogging. quickly”. • Ontology Maintenance and Publishing includes maintaining and publishing Table 1 ontologies, which might be done by Some proposed user groups and task types for dedicated information workers, i.e. the Semantic Web [10] ontologists, or by end-users themselves User Task Types Examples of tasks in Web 2.0 sites when they develop their Group vocabulary in an ad-hoc manner End users Information to look for a alongside indexing. seeking tasks restaurant near the Finally, Sabou et. al [12] propose the theater that will analysis of very specific Semantic Web still be open when the movie is over. applications from the point of view of users’ Information to organise the tasks. The kind of tasks under consideration in synthesis agenda of a this work, Ontology Matching, Folksonomy tasks conference Enrichment and Word Sense Disambiguation, attendant. are quite complex and can be decomposed in 50 simpler ones. Moreover, these tasks are point of view of an end-user, i.e. without targeted to users with some, or quite a lot, considering the particularities of the Semantic knowledge about the Semantic Web. Web. Some examples of particular end-users’ tasks are then presented, together with 3. Semantic Web End-User Tasks references to the related tasks in the literature previously analysed. Finally, the tasks are analysed deeper, considering what the From the analysis of the existing literature, Semantic Web might contribute to them, what Semantic Web applications and our experience technologies and methodologies make them with the Rhizomer platform [7], presented in possible and constitute and added value for Section ¡Error! No se encuentra el origen de them. la referencia., we have synthesized a set of All the considered tasks are basic ones. generic user tasks that can assist Semantic Web Usually, in the context of concrete Semantic developers while following a UCD approach. Web application, user tasks will be composed As it has been mentioned in the introduction, if of a mixture of these basic tasks. The objective this approach is followed, it is required to is to define a basic set of user tasks, that would define user tasks before the development facilitate tasks analysis and UCD while being process can continue. And prior to this, the easily combinable in order to derive more target user must be defined. complex and specific user tasks. In order to Though the objective is to define a set of illustrate these features, this categorisation of generic tasks, not specific to a particular tasks will be put into practice in order to analyse application, it is important to provide a minimal a generic platform for Semantic Web characterisation of the user and avoid defining applications in Section ¡Error! No se them as just any user. This is also motivated by encuentra el origen de la referencia., where the fact that the objective is to define the we anticipate that basic user tasks will. In the foundations for UCD of applications to be latter, the idea is that more specific user tasks adopted by as many users as possible. The most will be detected, which might be built from one populated user profile is end-users, that from or more of these basic Semantic Web user tasks. the point of view of the Semantic Web will in most cases stand for user with no or limited knowledge about the Semantic Web 3.1. Search technologies and methodologies [10]. One particularity is that we do not include in In general, this kind of tasks corresponds to this category domain experts with some those when a user poses a query and obtains a knowledge to be formalised as ontologies. They set of results that might be rendered in different might have limited knowledge about the ways. We include here when the search might Semantic Web particularities, but they might be delayed or repeated in the future, like in not be considered end-users because they have monitoring scenarios. very different objectives, and consequently they will carry out quite different tasks in order 3.1.1. Examples to accomplish them. The users we are considering are used to web and other Internet applications like Concrete examples of this task are when a electronic mail or instant messaging. Therefore, user performs a simple keyword-based search we should also consider a broader using a Web search engine, an advanced query categorisation of tasks users carry out online. It that allows constraining different search is necessary to consider a wider range of tasks dimensions, query by example, monitoring because, although the Semantic Web is not elections results or a sports match, etc. widely deployed right now, its opportunity is to underpin the whole range of online user 3.1.2. Semantic Web Search experiences and contribute new ways to do things in a more usable and accessible way. In the context of the Semantic Web, the user The next subsections present the set of can benefit from the implicit semantics when generic Semantic Web end-user tasks we performing a search and get more accurate propose. Each task is first considered from the results, i.e. higher precision and recall. 51 Moreover, the knowledge captured in music, viewing movie trailers, to follow a link ontologies can be used in order to guide the user received in an e-mail, etc. through the query construction process, in order to facilitate query by example or results 3.2.2. Semantic Web Browse presentation [13]. However, it is important to consider that most users are used to perform this kind of tasks In the context of the semantic web, it is by simple means like an input field where they possible to build a richer browsing experience type the keywords they are interested in. because the underlying model is built from Consequently, they might be confused if a more component of a smaller granularity, the triples sophisticated form or syntax is required in order formed by a subject, a predicate and an object. to pose a query. The combination of many triples builds up a In some cases, and after some user testing, it graph. This graph might be browsed by might be concluded that it is preferred to hide following the links between graph nodes all these subtleties from the user, to keep a following different criteria, not but just simple user interface, make use of the available showing the graph structure to the user [14]. semantics as part of the query engine internal For instance, those triples might come from mechanisms and exploit the semantics from the different “documents”. All the triples from a point of view of user interaction when document, identified by a URI might be presenting the results and as part of the displayed to the user, who can follow the links browsing user tasks, presented next. to external documents or browse the current data if it is not displayed all at once. An example of this behaviour is followed by 3.1.3. Related Work Tabulator [15]. It shows all the triples from a Semantic Web document as an unfoldable tree. This task includes Locating and Monitoring Another alternative is to provide a faceted (Heath), is similar to Fact Finding view if the metadata being browsed is and considers the temporal dimension of homogeneous, all the resources being browsed monitoring scenarios (Kellar). It also considers have similar properties describing them. This is Information Seeking (Battle) though some of possible using tools like Exhibit [16]. In the examples Battle et al. propose may require addition to the explicit metadata structure, it is other tasks, e.g. some sort of mash up in the also possible to take profit from the underlying case of combining news from different news ontologies in order to derive new links among sources. This task is also related to Semantic resources using mechanism like inference, Content Consumption (Mäkelä), though it is clustering [17] or semantic queries to other more specific because Semantic Content sources, for instance in order dynamically Consumption also includes when users browse suggest related products based on the semantic search results. Moreover, it is also a component description of the product being browsed. of the complex user task Word Sense Disambiguation (Sabou). 3.2.3. Related Work 3.2. Browse This task is related with both Exploring and Grazing (Heath), and also with Browsing This task is performed when the user moves (Kellar). Some of the examples of Information through the information currently displayed. In Seeking (Battle) also include aspects related the context of Web information systems this is with this task, e.g. learning more about a topic. usually done by following the links that connect This task is also related to Semantic Content the information to related information pieces. Consumption (Mäkelä), though it is more specific because Semantic Content 3.2.1. Examples Consumption also includes searching. Additionally, it is also a component of the complex user task Word Sense Disambiguation Concrete examples of this task are when a (Sabou). user gets informed about the latest news, reading blogs, entertainment, listening to 52 3.3. Annotate This task is connected with Asserting (Heath), especially if we consider that the statements made are metadata. It is also related In this task the user describes a resource by with a broader task that considers maintaining providing properties and values that model its information, Maintenance (Kellar), and also characteristics, its relations to other resources, with a more specific one that concentrates on etc. This task includes providing a completely updating content, Content Update (Battle). new description but also complementing an Considering tasks identified in the literature existing one, modifying it or deleting some or in the context of Semantic Web applications, all of the attributes currently available. this task is related with Content Indexing (Mäkelä) in the sense that by that task 3.3.1. Examples semantic annotations are generated, but just as long as some user intervention is required. Concrete examples of this task are when a Otherwise, it is not a user task but a system task. user tags a particular URL as it bookmarks it, It can be also related with Ontology providing the title and the description of a Maintenance and Publishing (Mäkelä), though video, geographically locating a photo, defining from the end-user characterisation we have a user profile that includes personal details and made this task lays outside the set of user tasks preferences, etc. under consideration. 3.3.2. Semantic Web Annotate 3.4. Mashup The main particularity of this task, in the This task is about the user gathering context of the Semantic Web, is that the different pieces of information and combining annotations are based on a formal model. them in order to get something more than the Consequently, annotations go beyond informal simple aggregation of those pieces. In other and ambiguous tags into properties and values word, the user tries to get something from the that might be constrained by the specifications aggregation of the data that cannot be or is captured in schemas and ontologies. This difficult to obtain from those pieces separately, feature is not just a way to facilitate machine without combining them into a coherent view. processing; it might be also as a way to facilitate the annotation task for the user. 3.4.1. Examples The user can benefit from a domain specification defining the available kinds of Concrete examples of this task range from resources, their properties depending on the simple mashups such as combining a set of resource type and the corresponding values. It resources that are geographically situated in is up to the user interface to guide the user order to, for instance, which are the hotels near through this knowledge space, dynamically a venue, or resources with temporal dimension constraining the choices to be made depending that are arranged in a calendar or timeline in on previous user actions, the context of use and order to facilitate scheduling. More the intended goals. complicated scenarios are also possible, which An example of a tool giving support to this are based on combining the sources of task in the context of the Semantic Web is the information without a predefined output view, Semantic Forms extension2 for Semantic like combining a local list of publications with MediaWiki [18], which takes profit from the information about their impact factor in order to underlying semantic models that structure compute the overall impact, detect trends, available types, properties and their values. highlight the more relevant publications, Tabulator also has recently introduced some preparing a research activity report, etc. support for metadata edition [19]. 3.3.3. Related work 3.4.2. Semantic Web Mashup 2 http://www.mediawiki.org/wiki/Extension:Semantic_Forms 53 In the context of the Semantic Web, this task might be used in descriptions for many involves combining two or more pieces of resources, some of which the user might not be metadata about common resources in order to aware of it at the moment. aggregate the available descriptions about Results from a mapping task might be used them. It is also possible that the metadata is in order to facilitate or automate a mashup, or about different resources, but in this case, they both tasks might be carried out alternatively and should be similar in some sense in order to co-ordinately as a process where the user is make possible the aggregation in some mashing up a set of resource descriptions and dimension, e.g. they all have geographical during that process some mappings among the coordinates or are situated in time and can be vocabularies being used are defined. placed together in a map or timeline It might be the case that mappings are respectively. derived from the analysis of the interaction of The main benefit of Semantic Web many users, however this is a system task, the technologies and methodologies for this task is user does not directly and consciously intervene that as semantic metadata and ontologies are in this case. Here we are referring to tasks available, it is easier to implement some sort of initiated by the user. assistance for the user during the aggregation process. The assistance may range from the 3.5.1. Examples ability to propagate the aggregations made to one particular resource property to all the uses of that property in the metadata being mashed Examples of this task range from simple up, like in the Potluck mashup tool [20], or scenarios like stating that two tags are equivalent to more complex ones like relating exploiting in a more automatic way the different product categories or stating that all available semantic metadata using semantic and the things that one repository classifies as statistical measures in order to provide a papers are also a kind of publication as preliminary mashup that the user might then specified in a second repository vocabulary. customise, like in the case of the semantic information mashup tool Sig.ma [21]. 3.5.2. Semantic Web Map 3.4.3. Related work In the context of the Semantic Web, and also considering that we have characterised all these This task is related with, though slightly tasks as those for an end-user, this task more specific than, Information Gathering corresponds to when the user defines simple (Kellar) and includes the main characteristics of mappings among classes, properties and values both Evaluating and Arranging (Heath), which specified in different ontologies. It is not about might also involve search and browse but exhaustive mappings among ontologies but whose added value is about combining instead about specific mappings that might be information and extracting something more that usually justified by the need of facilitating its pure addition. It is also related with mashing up some resource descriptions, or Information Synthesis (Battle) and the Semantic making the mashup more systematic. Content Consumption user task (Mäkelä), which is much wider and also includes searching and browsing. 3.5.3. Related work 3.5. Map This task is a particular case of Ontology Mapping (Battle), geared towards very simple mappings and usually triggered by the system This task takes place when the user defines that asks users for confirmation because we mappings among terms from different focus on end-users that are not ontologists. It is vocabularies. It is not constrained to a particular also related with Ontology Maintenance and set of resources like in the case of the Mashup Publishing (Mäkelä) but that is also a task task, and it does not operate at the level of geared toward ontologists and domain experts, particular resource descriptions. On the not end-users. The same applies when contrary, in this task, the user is working at the considering Content Indexing (Mäkelä). The level of the vocabularies. These vocabularies 54 same applies for Ontology Matching (Sabou) Annotate as the user does not perceive as being while Folksonomy Enrichment might be easier describing something. and more appropriate for end-users. 3.6.3. Related work 3.6. Share This task is similar to Sharing (Heath) and This task considers uploading, publishing, Information Sharing (Battle). There are not updating and deleting pieces of content with the tasks related with this one in the studies of intention of making them available to other Semantic Web tasks we have analysed. This users, who can access the content from a place seems related with the fact that, as we have and at a time individually chosen by them. This previously said, in the context of the Semantic last statement allows to clearly distinguishing Web sharing semantic descriptions is a task this task from the Communicate task, which is included in annotation tasks. presented next. This task is also differentiated from 3.7. Communicate Annotate in the sense that what is added, edited or removed is not metadata, data about data, but data itself. This data will usually correspond to This task is about sharing information different kinds of content that users want to directly with particular users, without the share online, like videos, text, images, etc. intention of making it available to other users from a place and at a time individually chosen by them. The process is in this case driven by 3.6.1. Examples the user participating in this task as the emitter. Examples of this task are posting a blog or 3.7.1. Examples micro-blogging, participating in a forum, sharing a photo in a social network, making a file available through a Peer-to-Peer network, Examples of this task are to participate in a etc. The participation in forums might also be chat, to send an e-mail, video-conference, etc. seen as a communication task, described below, We have included here e-mails because they are but the intention in most forums is to build a usually kept private and not intended to make piece of information around a subject and to them publicly available to other users apart make it available for later use. from the recipient. Moreover, the communication is driven from the emitter as long as the recipient has the e-mail client up and 3.6.2. Semantic Web Share listening. In the context of the Semantic Web, this 3.7.2. Semantic Web Communicate task, as long as related with data and not with metadata, is not directly supported by Semantic Web technologies and methodologies. Thought communication management is However, it might be enriched by triggering fundamentally related with other tasks, like some sort of content indexing and automatic searching for a specific e-mail, browsing metadata generation. The metadata just conversations or annotating an e-mail, there generated can then trigger an Annotate task, might be room for semantic technologies to which allows the user editing and managing this play a role during this task. In any case, this is metadata. one of the areas with less results coming from However, there might be also scenarios the Semantic Web, and also the one that seems where the distinction between data and to provide less room for them. metadata is somehow blurred. For instance, in For instance, Haystack [23] is a tool for the the context of Linked Data [22] publishing, web and desktop that helps the user manage where bunches of semantic data are made whatever information a user considers important, which includes communications available without considering the specific based on e-mail. The e-mail is processed and resources being described, the task from the semantically annotated in order to perform point of view of the user is a Share, not an 55 communication management based on these place. For instance, by facilitating form filling semantic annotations. while the user provides the required data to complete the transaction. Another way to 3.7.3. Related Work support the transaction might be adapting the results to user preferences and context, for instance performing currency conversions This task is related with Notifying and following user preferences. Discussing (Heath). 3.8.3. Related Work 3.8. Transact This task is present in two of the tasks lists This is the last task, it is associated with user considered. There are Transacting (Heath) and actions that provoke a change in the state of a Action-oriented (Battle) tasks. real-world entity or of a resource in a system outside the scope of the system the user is interacting with. 4. The Rhizomer Platform Testbed 3.8.1. Examples Rhizomer3 is a platform based on Semantic Web technologies that facilitates publishing semantic data and building interactive Semantic Examples of this task are buying a book, Web applications on top of it. Rhizomer differs ordering a money transfer between bank from semantic web browsers in the sense that it accounts, etc. The range of specific tasks is not just a browser application; there is also a included in this category might vary a lot server part that allows defining datasets to be depending on the interactive system attention is explored and which interacts with the SPARQL focused on. If we concentrate on the user tasks endpoints holding the datasets semantic data. for a specific application, any task that involves However, it is also capable of browsing data not interacting with other systems might be stored in SPARQL endpoints but linked from considered a transaction as a way to focus the them. analysis. For instance, if some resource from On the other hand, if a broader system is DBPedia [24] is used in a description stored in considered, for instance any information dataset published through Rhizomer, it is system, the study might be detailed further and possible to retrieve the associated metadata by particular tasks among the ones presented following the Linked Data principles and to before might be identified as the goal of that perform all the user tasks available for local interaction. In any case, actions that take place data in a way totally transparent for the user. in the real world, outside interactive Moreover, Rhizomer also provides information systems, might be considered mechanisms that facilitate integrating external transactions in the context of this end-user tasks web services in a dynamic way. The external proposal. services to be integrated should be semantically described and those descriptions should specify 3.8.2. Semantic Web Transact the kind of resources (classes) the get as input. Rhizomer implements the mechanisms that Together with communicate, this is the task allow associating at run time the resources that might be less influenced by Semantic Web classified as being of the input type with the technologies and methodologies. This is due to corresponding service. the fact that by the definition of this task, they These associations are implemented as links correspond to interactions of the user with that allow the user invoking the service, that systems outside the Semantic Web. will receive the resource description as input, In any case, applications might take profit process it and return some output. Usually, this from these technologies and methodologies output will be HTML content to be integrated while supporting this task before and after the into the interface. This way, it is easy to plug in processing outside the Semantic Web takes external services that provide bridges to other 3 Rhizomer, https://rhizomer.rhizomik.net 56 services but also new ways to interact with • Mashup: mix two or more pieces of resources. metadata about common resources, or The backend is based on a web application resources similar in some sense, e.g. providing an API to interact with the defined they all have geographical coordinates datasets and the SPARQL endpoint that or are situated in time and can be placed constitute them. The backend is built on top of together in a map or timeline previous API and implemented using a web respectively. framework based on HTML and JavaScript that • Map: define simple mappings between makes the user interface highly interactive. concepts from different ontologies. Rhizomer gives support to most of the users’ • Share: upload, update and delete tasks presented in the previous section, with the pieces of content (HTML, images, exception of the Communicate task: videos, etc.). • Transact: generically, this task • Search: pose semantic queries using includes any user action that change the HTML forms, which are dynamically state of a real-world entity or of a generated and obtain resource resource in a system outside Rhizomer. descriptions rendered as HTML, as shown in Figure 1. 5. Conclusions and Future Work • Browse: navigate through the graph of data retrieving fragments of manageable size and rendering them as In order to make the Semantic Web reach interactive HTML, as shown in Figure “real world” end-users, special care must be 2. placed in making Semantic Web applications more usable and accessible. It is possible to take • Annotate: provide new semantic profit from the experience accumulated by the metadata describing a resource, or edit Human-Computer Interaction community and existing one, using HTML forms that apply User-Centred Design approaches. assist the user during this process. Figure 1. Rhizomer’s faceted view featuring search 57 Figure 2. Rhizomer browsing Linked Data Web end-user tasks proposed is based on the This kind of approaches place the user at the analysis of existing tasks inventories for the centre of the development process and start by Web and even for online information systems defining the user, its context and the tasks to be in general. performed by them in order to meet their needs. The analysis is complemented with the These tasks are specific to a particular experience gained implementing a Semantic interactive application, but it very useful to Web platform and a Semantic Web application define a set of common user tasks in the context based on that platform. Additionally, a list of of a particular domain, e.g. Web information Semantic Web capabilities has been used in systems, in order to establish UCD guidelines order to complement the process of checking and common interaction patterns for that the consistency and coverage of proposed set of domain. end-user tasks. This is the main aim of this work, to identify The set of tasks includes Search, Browse, a set of common user tasks for the Semantic Annotate, Mashup, Map, Share, Communicate Web. However, it is important to concretise the and Transact. Each of these tasks has been user profile, it is not enough to say that tasks are described avoiding technological for any user. In this case, as the aim is to considerations and then presented from the contribute to the widespread adoption of the point of view of the Semantic Web. They are Semantic Web, the target user is the end-user. also related to the tasks proposed in the This is a user with no or quite limited literature under consideration. knowledge about the Semantic Web. The Finally, the set of tasks has been put into context is any online application based on practice and the Rhizomer platform in order to Semantic Web technologies. detect if they give support to these end-user Once the user and the context are defined, it tasks and that not additional ones are required. is time to determine the user tasks. In order to This analysis also allows, in the case of the consider a broad range of user tasks, it is Rhizomer platform, presenting how these tasks important to avoid constraining the analysis to are materialised in the context of the platform the current Semantic Web. The set of Semantic as different interaction patterns. 58 Additionally, it is shown how, as it might be a user tasks, once decomposed into its basic anticipated, the platform user tasks are the basic components. ones while for the application build on top of For instance, Welie proposes patterns for Rhizomer, the more complex user tasks can be Browsing or Searching. The objective is to decomposed into basic user tasks from the build on top of these pattern libraries and proposed set. classify them considering the proposed Future work concentrates now on the next Semantic Web end-user tasks. Many of them natural step when following a UCD approach. might be mapped directly from the Web domain Once the user tasks have been identified, it is to the Semantic Web, like Welie’s patterns for really useful to have an inventory of interaction Search and Browse, two tasks near clear patterns that give support to these user tasks as equivalents in the Semantic Web. In any case, a guideline. There are many lists of interaction they must be studied in detail, new patterns, though most of them focus on Web opportunities should be detected and there is systems or other interactive systems without also room for novel interaction patterns that the particularising their proposal in the context of Semantic Web might make possible. On the Semantic Web applications and Semantic Web other hand, our analysis against related work user tasks. specific for Web information systems shows Our aim is to build an inventory of Semantic that the user tasks where the contribution of Web interaction patterns starting from existing Semantic Web technologies might be more inventories, e.g. Tidwell’s [25], van Welie’s important, because they a less consider or not [26], Toxboe’s [27] or Crumlish & Malone's considered at all, are Annotate, Mashup and [28]. Some of them are structured in part using Map. user tasks as the way of classifying the patterns, making it possible to use them as a reference 6. References when implementing the interaction that support [1] M.D. Lytras, R. García. Semantic Web 2013, pp. 1025-1045. DOI: 10.3217/jucs- applications: a framework for industry and 019-08 business exploitation–What is needed for [6] R. García, R. Gil, E. Bakke, D.R. Karger. the adoption of the Semantic Web from the A benchmark for end-user structured data market and industry. International Journal exploration and search user interfaces. of Knowledge and Learning, 4(1), 2008, Journal of Web Semantics, 65, 2020. DOI: pp. 93-108. DOI: 10.1016/j.websem.2020.100610 10.1504/IJKL.2008.019739 [7] J.M. Brunetti, R. García. User-centered [2] T. Heath, J. Domingue and P. Shabajee: design and evaluation of overview "User interaction and uptake challenges to components for semantic data exploration. successfully deploying Semantic Web Aslib Proceedings, 2014, 66(5), 2014, pp. technologies". In Proc. 3rd International 519-536. DOI: 10.1108/AJIM-12-2013- Semantic Web User Interaction 0153 Workshop, Athens, Georgia, USA, 2006. [8] T. Heath, M. Dzbor, E. Motta: "Supporting [3] m. c. schraefel, J. Golbeck, D. Degler, A. User Tasks and Context: Challenges for Bernstein, L. Rutledge, Semantic web user Semantic Web Research", in: Proc. interactions: exploring hci challenges, in: ESWC2005 Workshop on End-User CHI Extended Abstracts, 2008, pp. 3929- Aspects of the Semantic Web 3932. (UserSWeb), 2005. [4] ISO/IEC 25010-3. Systems and software [9] M. Kellar, C. Watters, M. Shepherd. A engineering: Software product quality and Goal-based Classification of Web system quality in use models. ISO Press, Information Tasks. In 69th Annual 2011. Meeting of the American Society for [5] J.L. González-Sánchez, R. García, J.M. Information Science and Technology, Brunetti, R. Gil, J.M. Gimeno. Using Austin, US, 3-8 November 2006. SWET-QUM to compare the quality in use [10] L. Battle, Preliminary Analysis of Users of semantic web exploration tools. Journal and Tasks for the Semantic Web, Semantic of Universal Computer Science, 19(8), Web User Interaction Workshop, 59 [18] M. Krötzsch, D. Vrandečić, M. Völkel, International Semantic Web Conference, Semantic MediaWiki, in: The Semantic Athens, GA, 2006. Web - ISWC 2006, Heidelberg, DE, [11] E. Mäkelä, K. Viljanen, O. Alm, J. Springer, 2006, pp. 935-942. Tuominen, O. Valkeapää, T. Kauppinen, J. [19] T. Berners-Lee, J. Hollenbach, K. Lu, J. Kurki, R. Sinkkilä, T. Känsälä, R. Presbrey, E. Pru d'ommeaux, M.C. Lindroos, O. Suominen, T. Ruotsalo and schraefel, Tabulator Redux: Writing Into E. Hyvönen. Enabling the Semantic Web the Semantic Web, Southampton, UK, with Ready-to-Use Web Widgets. FIRST - Electronics and Computer Science, First Industrial Results of Semantic University of Southampton, 2007. Technologies Workshop, at the [20] D. Huynh, R. Miller, Robert, and D.R. ISWC+ASWC Conference, Busan, Korea, Karger, Potluck: Data Mash-Up Tool for November 11-15, 2007. Casual Users, 6th International Semantic [12] M. Sabou, J. Garcia, S. Angeletou, M. Web Conference (ISWC), November D'Aquin and E. Motta. Evaluating the 2007. Semantic Web: A Task-based Approach” [21] M. Catasta, R. Cyganiak, G. Tummarello, in The 6th International Semantic Web Towards ECSSE: live Web of Data search Conference and the 2nd Asian Semantic and integration, Semantic Search 2009 Web Conference, 2007. Workshop, Madrid, Spain, 2009. [13] M.L. Wilson, m.c. schraefel, R.W. White, [22] C. Bizer, T. Heath, K. Idehen, T. Berners- Evaluating Advanced Search Interfaces Lee, Linked data on the web using Established Information-Seeking (LDOW2008), en: Proceeding of the 17th Models, Journal of the American Society International Conference on World Wide for Information Science and Technology, Web, Beijing, China, ACM, 2008: págs. (2009) in press. 1265-1266. [14] D.R. Karger, m.c. schraefel, The Pathetic [23] D. Quan, D. Huynh, D. Karger, Haystack: Fallacy of RDF, SWUI 2006 - 3rd A Platform for Authoring End User International Semantic Web User Semantic Web Applications, en: The Interaction Workshop, November 2006. SemanticWeb - ISWC 2003, LNCS Vol [15] T. Berners-Lee et. al., Tabulator: 2870, Springer, 2003, pp. 738-753. Exploring and Analyzing linked data on [24] J. Lehmann, et al. DBpedia–a large-scale, the Semantic Web, in: Procedings of the multilingual knowledge base extracted The 3rd International Semantic Web User from Wikipedia. Semantic web, 6(2), Interaction Workshop, SWUI’06, Athens, 2015, pp. 167-195. DOI: 10.3233/SW- Georgia, 2006. 140134 [16] D.F. Huynh, D.R. Karger, R.C. Miller, [25] J. Tidwell, Designing Interfaces: Patterns Exhibit: lightweight structured data for Effective Interaction Design, O'Reilly, publishing, en: Proceedings of the 16th 2005. International Conference on World Wide [26] M. Welie. A Pattern Library for Interaction Web, Banff, Alberta, Canada, ACM, 2007, Design. Available from: pp. 737-746. http://www.welie.com [17] L. Rutledge, J. van Ossenbruggen, L. [27] A. Toxboe. User Interface Design patterns. Hardman, Making RDF Presentable - Available from: http://ui-patterns.com Global and Local Semantic Web [28] C. Crumlish, E. Malone. Designing social Browsing, in: Proceedings of The interfaces: Principles, patterns, and Fourteenth International World Wide Web practices for improving the user Conference, WWW’05,ACM Press, experience. O'Reilly Media, 2009. Chiba, Japan, 2005, pp. 199-206.