Artificial Intelligence (AI) is receiving significant attention in various domains, such as precision medicine, mental healthcare, manufacturing, education, finance, and many others. Most applications in these domains are concerned with the classification or language generation tasks, wherein confidence in the outcome is measured through traditional metrics and stakeholder agreement. When 91% of companies in these domains started a call for explainability in AI, it propelled the infusion of Knowledge Graphs (KGs) to make AI explainable. Further, the synthesis of AI and KGs added new capabilities in data-driven AI, such as (a) improved performance with minimal training data, (b) user-level explainability, (c) modeling rare events and handling uncertainty, (d) inducing context sensitivity in AI, and (e) better control over the behavior of AI system to ensure safety. Furthermore, with AI and KG together, personalization became possible. The induction of Personal Knowledge Graphs or Personalized Knowledge Graphs (PKGs) began to endure the capabilities of AI that could customize the outcome based on users' persona and enhance user engagement. Moreover, with PKGs, the community of AI could implement trustworthy systems safeguarding user security and ensuring consistency and robustness in outcomes. This chapter allows interdisciplinary researchers and practitioners in AI to contribute their research on methodologies, tools, and applications that discuss PKG's construction, utilization, and inference functionality.

A blog post at Google on 16 May 2012 with the title "Introducing the Knowledge Graph: things, not strings" represented one of the first references concerning the current definition of Knowledge Graphs. On the other hand, the article by Krisztian Balog et al. "Personal Knowledge Graphs: A Research Agenda" represented one of the very first references clearly defining the commonalities of Personal Knowledge Graphs (PKGs) such as presenting a "Spider-Web" graph layout having as the user its "center of gravity." To date, the literature related to PKGs is currently scarce given that it is still a virgin and promising research field. In this chapter, we present a survey including a classification of different types of applications of PKGs, spanning from E-learning Systems to Personal Information Managers (PIMs), to the Decentralized Web (e.g. the "Social Linked Data" (SOLID) stack), and so on. This classification identifies nine overlapping categories given that PKGs may belong to one or more categories. In each classification, we focus/highlight common and outstanding architectural components as reference architectures for each category type. We end-up the chapter by including and suggesting a reference architecture depicting desired main components for a semantic web (SW)-based PKG.

In this chapter, we have described the semantic web stack, a comprehensive list of open standards for representing, reasoning with, querying and validating knowledge graphs. We have described different projects relying on those standards to build personal knowledge graphs, and to allow them to seamlessly interoperate with other knowledge graphs on the web. Finally, we discussed related standards for describing rules and policies, in order to express under which conditions information from personal knowledge graphs can be shared with others, in the advent of the so-called data economy.

The digitization of our lives creates new complex social models. Raymond [48] described how Internet and the web created the conditions without which the open-source ecosystem could not have developed. Creating the conditions of a secure and privacy-preserving data economy in an ever more interconnected world remains an ongoing and challenging endeavor.

Work on personal knowledge graphs would greatly benefit from decades of progress in the cognitive sciences. Knowledge is about understanding information based upon past experience, where understanding enables reasoning and decision-making. Today's applications, however, embed limited understanding in application code, as deductive logic is not adequate for human-like reasoning. Humans are always learning and never attain perfect knowledge, and our reasoning out of necessity has to deal with uncertain, incomplete, imprecise, and inconsistent knowledge. This chapter will introduce a cognitive approach to personal knowledge graphs, including plausible reasoning, System 1 for intuitive thinking, including effortlessly and rapidly generating coherent explanations, e.g., for natural language understanding, and System 2 for effortful slower deliberative thought, and how this can enable human-like memory, reasoning and decision-making.

Entity resolution (ER) is the problem of determining when two entities refer to the same underlying entity. The problem has been studied for over 50 years, and most recently, has taken on new importance in an era of large, heterogeneous 'knowledge graphs' published on the Web and used widely in domains as wide ranging as social media, e-commerce and search. This chapter will discuss the specific problem of named ER in the context of personal knowledge graphs (PKGs). We begin with a formal definition of the problem, and the components necessary for doing high-quality and efficient ER. We also discuss some challenges that are expected to arise for Web-scale data. Next, we provide a brief literature review, with a special focus on how existing techniques can potentially apply to PKGs. We conclude the chapter by covering some applications, as well as promising directions for future research.

Personal knowledge graphs (PKGs) refer to knowledge graphs (KGs) that contain information about personal descriptions. Thus, PKGs contain definitions and/or descriptions of information concerning a person or user. The development of PKGs is an essential aspect of decision-making applications where it may be necessary to involve user information such as preferences (tastes, styles, colors), physical characteristics (health, ergonomic), and historical (academic, legal), to mention a few. One of the main aspects of constructing such structures lies in the information components connected to form a graph. These components, called triples, are composed of Subject-Predicate-Object elements, where the Subject is oriented to be a real-world entity about a person, the Object is an entity or value that is said of the Subject, and the Predicate is the relationship that exists between these two things. However, the construction of PKGs is often a challenge, whether performed manually (by an expert) or automatically (computationally). On the one hand, their definition has not been widely studied, so the construction of PKGs is time-consuming even if performed manually. On the other hand, their computational construction is a process that, according to the source of information (and its structure), can implicate various language challenges to obtain the necessary triple elements accurately (to extract the relations that will be used in the construction and connection of triples). This chapter presents an overview of techniques for extracting relations applied in the construction of PKGs. The idea is to provide conceptual aspects on the extraction of relations for the construction of KGs and then to review the most common techniques applied to PKGs. Finally, some considerations in the extraction of relations and the construction of PKGs are presented.

Personal knowledge graphs (PKGs) deal with user-centric information and focus on personalizing data from various sources of interest to the user. They help to organize information deliberately and improve the competencies of both the individual and the organization. Visualizing knowledge graphs also enhances the users' understanding of the data and expands users' decision-making capabilities. This chapter presents the concept of PKGs and their visualization tools. We will discuss the benefits of knowledge graphs and compare knowledge graphs with PKGs. The main advantages and applications of PKGs are also illustrated in this chapter. Moreover, various personal knowledge management and visualization tools like Obsidian, RemNote, TiddlyRoam, Dendron, Logseq, Foam, and Roam Research are also presented here.

Query-answering (QA) is one of the key areas in Artificial Intelligence, where various researches are performed in recent years. Building query-answering system helps the organization of all sectors. Generating automatic responses saves both time and money. We examine the problem of query-answering over knowledge graphs (KG) where various QA approaches focus on simpler queries and do not work very well for complex queries or vice versa. In addition to that, reasoning over KG is also to be handled properly to predict the proper answer to the corresponding query. Models that use SPARQL are good at domain-related queries, but they are unable to handle out-of-domain queries. Combining contextual text representation and semantic graph representation is a challenge. Our area of research is to combine text and KG for open domain query-answering. Adapting the joint representation ensures that the model can perform well in both simple and complex queries. In this chapter, we explain the various works that have been conducted and the challenges that have come along with it.

Personal knowledge Max Planck Institute for Informatics, Germany graph (PKG) is an extremely useful asset, supporting many downstream applications, which facilitate user experiences. However, populating the PKG with the facts about the user is time-consuming and requires continuous revisions for the changeable user attributes. Instead of outsourcing this task to humans, one can leverage the existing machine learning methods to automatically populate a PKG. In this setting, personal facts can be extracted from the user's conversational data, such as chats and social media submissions. This kind of data exists in abundance for almost every person and contains rich signals of personal attributes.

In this chapter, we provide a comprehensive survey of existing methods for personal knowledge extraction from conversational data, including inferring demographic facts, interests and relationships of the speakers. We discuss the pros and cons of the current approaches and propose directions for future work.

Knowledge graphs (KGs) can Aarhus University, Denmark be leveraged to effectively answer complex questions involving relationships among entities. Such question answering has important applications in various scientific areas such as biology, medicine, and engineering. Often the graphs occurring in these domains are massive consisting of many millions or billions of facts, which poses challenges for the methods that are used for question answering. In this chapter, we focus on one promising approach for handling the massive-size challenge, namely fact summarization for personalized knowledge graphs, which is a variant of graph summarization.

The importance of personalized knowledge graphs (PKGs) cannot be understated as they are applied across various spheres of our lives where they continuously had a positive impact. PKGs are a form of knowledge graphs that have completely changed the way item recommendations are made by introducing explainable recommendations, and alleviating the cold-start and data sparsity issues which are common in the traditional recommender systems. These heterogeneous graphs contain connectivity patterns and semantic relations which can be exploited for quality recommendations. This chapter focuses on explaining how PKGs can be used to recommend items to users when faced with a large variety of items to select from. Methods can be used for the implementation of accurate and explainable recommendations via PKGs such as embeddings and structure-based methods and are presented as well as the challenges and limitations entailed in applications development using this technology. The application of PKG recommendations in commerce and finance, Internet of Things (IoT), and healthcare are used to demonstrate its utility and competitive performance. In order to appreciate the importance of PKG recommendation systems, an example application of PKG recommendation in the field of personalized badminton training is also presented.

Knowledge graphs (KGs) provide structured data for users' applications such as recommendation systems, personal assistants, and question-answering systems. The quality of the underlying applications relies deeply on the quality of the knowledge graph. However, KGs inevitably have inconsistencies, such as duplicates, wrong assertions, and missing values. The presence of such issues may compromise the outcome of business intelligence applications. Hence, it is crucial and necessary to explore efficient and effective methods for tackling the evaluation of KGs. These techniques get much tougher when the KG deals with personal data, which is referred to as Personal Knowledge Graph (PKG). This chapter covers PKG's creation, population, and more importantly their evaluation from a data quality perspective.

The advent and the rise of the Internet of Medical Things (IoMT) has led to the proliferation of wearable devices that collect a vast amount of personal health data. This data can create personal health knowledge graphs to help individuals better understand their health, manage chronic conditions, and improve overall well-being. This chapter will explore the process of collecting data from wearable devices, creating personal health knowledge graphs, and utilizing them in various health applications. Such personal health knowledge graphs are useful for healthcare providers in assessing health outcomes of patients that are not evident to them during the clinical visits. However, not all the data will be useful to healthcare providers, and further analysis is needed to make sense of the data, for which semantic technologies would help. We will outline how to extract data from personal health devices, transform the data into an appropriate modality, and load it to relevant systems for consumption by healthcare providers.

This chapter covers the topic of integrating personal knowledge graphs (PKGs) into a large enterprise. When the term refers to a personal knowledge graph (PKG) that is integrated into the enterprise, it is used as integrated personal knowledge graph (IPKG). The proposed chapter consists of four parts: Background on PKGs to explore about the fundamentals of PKGs; IPKG challenges to explore the challenges of integrating PKGs into an enterprise knowledge graph and the role that machine learning can play in the future; IPKG steps to show the flow of steps in performing the integration; Security to find grain access control - using role and authorization.

Personal knowledge graphs (PKGs) are important for individuals, organizations, and industries to manage and utilize their knowledge effectively. With the increasing volume and complexity of information available today, PKGs can help individuals and organizations make sense of this information and extract meaningful insights.