Https://arxiv. Org/abs/2003. 02320
In this paper, we present a comprehensive picture of knowledge mapping, which has recently attracted considerable attention from industry and academia in the context of the need to develop diversified, dynamic, large-scale data collection. Following an overview, we summarised and compared the various map-based data models and query languages used for knowledge mapping. We'll discuss schema, education, and coNtext's role in knowledge mapping. We explain how knowledge is expressed and extracted using a combination of performing and summing techniques. We have summarized the methods of creating, enriching, assessing, fine-tuning and disseminating knowledge maps. We will outline the well-known open knowledge mapping and enterprise knowledge mapping and their application, and how they use the above-mentioned technology. Finally, we summarize the future directions of high-level knowledge mapping。
Advantages of knowledge mapping
Although the term “knowledge mapping” has existed in the literature since at least 1972, its modern form originated from google knowledge mapping published in 2012, followed by airbnb, amazon, ebay, facebook, ibm, lInkedin, microsoft, step up, etc. Have issued successive announcements to develop knowledge maps. It has proved difficult for academia to ignore the growing prevalence of the concept: an increasing number of scientific literatures publish themes on knowledge mapping, including books (e. G.) and papers outlining definitions (e. G.) and new technologies (e. G.
298, 399, 521
And surveys of specific aspects of knowledge mapping (e. G.)。
At the heart of all these developments is the idea of using graphics to represent data, which is often reinforced by some form of explicit expression of knowledge. The results are most frequently used for applications involving large-scale integration, management and extraction of value from different data sources. In this case, the use of map-based knowledge abstraction has many advantages compared to relationship models or nosql alternatives. The figure provides a concise and intuitive abstraction of various fields, where social data, biological interactions, bibliographic references and cooperative authors, and the (potential cycle) relationship between entities inherent in transport networks are captured. The figure allows the maintenance to postpone the definition of the model and allows the data (and its scope) to develop in a more flexible manner than is usually possible in relationship settings, especially with regard to obtaining incomplete knowledge. Unlike the (other) nosql model, a specific graphical query language supports not only the standard relationship operators (connection, association, projection, etc.), but also the navigation operators that are retrieving the entities connected by any length path. Standard knowledge indicates formalism - as it is
66, 228, 344
And rules - may be used to define and reason the semantics used to mark and describe the terms at nodes and sides in the diagram. Scalable graphical analysis framework
314, 478, 529
It can be used to calculate centrality, clusters, summaries, etc. To obtain insight into the areas described. Various forms of representation have also been developed to support the direct application of machine learning techniques on maps。
In summary, the decision to construct and use knowledge maps provides a range of techniques for integrating and extracting value from different data sources. However, we have not yet seen a common uniform summary describing how knowledge maps are used, what technologies are used and how they are relevant to existing data management themes。
Course objective: a comprehensive presentation of knowledge maps spectrum
The objectives of the academy are to provide a comprehensive picture of the knowledge maps: describe their basic data models and how to search for them; discuss the signs associated with schema, education, and context; discuss the way in which the simulations and aggregations are used to clarify knowledge; describe the technologies that can be used to create and enrich graphic structure data; describe how the quality of knowledge maps is identified and how knowledge maps are improved; discuss the criteria and best practices for publishing knowledge maps; and provide an overview of existing knowledge profiles found in practice. Our target audience includes researchers and practitioners who are not familiar with knowledge maps. We therefore do not assume that the reader has specific expertise in knowledge mapping。
Knowledge maps. The definition of “knowledge mapping” remains controversial
36, 53, 136
Some (sometimes conflicting) definitions have emerged, ranging from specific technical recommendations to more inclusive general recommendations; these earlier definitions are discussed in appendix a. Here, we have adopted an inclusive definition, in which we view the knowledge map as a data map designed to accumulate and transmit knowledge of the real world, the nodes of which express interest in entities whose margins represent the relationship between those entities. Data graphs (also known as data graphs) are compatible with a map-based data model, which can be a map with a bordermark, a attribute graph, etc. (we discuss specific alternatives in section ii). This knowledge can be accumulated from external resources or extracted from the knowledge map itself. Knowledge can consist of simple words such as “santiago is chile's capital” or quantitative words such as “all capitals are cities”. A simple statement can accumulate as a side of a data map. If the knowledge map is intended to accumulate quantitative statements, then there is a need for a more representative way of expressing knowledge — for example, the body or the rules. The evolutionary approach can be used to inherit and accumulate further knowledge (e. G. “santiago is a city”). Additional knowledge based on simple or quantitative statements can also be extracted and accumulated from knowledge maps by way of aggregation。
Knowledge mapping usually comes from multiple sources and may therefore be very diverse in terms of structure and particle size. Addressing this diversity, indicating patterns, identities, and context often plays a key role in defining in one mode a high-level structural knowledge map that indicates which nodes (or external sources) refer to the same real entity, while the context may indicate that the knowledge of a particular unit setting up some units is real. As noted earlier, knowledge mapping requires effective extraction, enrichment, quality assessment and refinement methods to grow and improve over time。
In practice, the goal of the knowledge mapping is to serve as the evolving knowledge base for sharing within the organization or community. In practice, we distinguish between two types of knowledge mapping: open knowledge mapping and enterprise knowledge mapping. Open knowledge maps are posted online to the benefit of the public. Most prominent examples - dbpedia, freebCase, wikidata, yago etc. - cover many areas, either extracted from wikipedia or built by volunteer communities. Open knowledge maps have also been published in specific areas such as media, government, geography, tourism
11, 263, 308, 540
Life sciences, etc. Enterprise knowledge mapping is usually intra-firm and applied to business examples. Prominent industries using enterprise knowledge maps include web search (e. G. Bing, google), business (e. G. Airbnb, amazon)
127, 280
(e. G. Facebook, l)In this context, i think, finance (e. G. Esenzel, bank of italy, bloomberg, capital one, bank of rich countries). Applications include search, recommendations
83, 205, 214, 365
Personal agents, advertising, business analysis, risk assessment, automation and more. In section 10 we will provide more details on the use of knowledge maps in practice。
Section 2 of the structure of the rest of the course provides an overview of graphical data models and the languages available for searching them. Section 3 describes the pattern, marking and context of the knowledge map. Section iv describes the evolutionary formalism through which knowledge can be described and derived. Section 5 describes summary techniques that can extract additional knowledge. Section 6 discusses how to create and enrich knowledge maps from external resources. Section 7 lists the qualitative dimensions that can be used to assess knowledge maps. Section 8 discusses technologies for the refinement of knowledge mapping. Section 9 discusses principles and protocols for the publication of knowledge maps. Section 10 presents some well-known knowledge maps and their applications. Section 11 summarizes the research profile of knowledge mapping and future directions of research. Appendix a provides the historical background of the knowledge map and previous definitions. The official definitions to be cited in the body of the paper are listed in appendix b。
