This page provide the background information about Project ReKisstory
ReKisstory Project is a project of Go Sugimoto (PhD candidate at Vrije Universiteit Amsterdam) with a mission to:
To achieve these goals, he has created a simple user-friendly search engine tool "ReKisstory". It allows users to search a large-scale database of human knowledge and analyze a broad spectrum of data with a special focus on time and space. Your search results will be presented in nice interactive visualizations (table, graph, timeline, and map (and network)). But, this is just a building block of what we envision. If you are interested in a bigger vision, please check out our blog post
It is a research project in a Computer Science department. Therefore, it is meant to be technically-oriented, and we expect technical audience in terms of cutting-edge technology, called Semantic Web. However, the author belongs to the User Centric Data Science (UCDS) group, meaning he would like to address a wide range of end-users, because there are many user aspects in this project. In addition, he works in the interdisciplinary filed, especially Digital Humanities (DH). So, he welcomes anybody who is interested in history, humanities, technology and science, and beyond. Still, potential target groups would be:
The author started the project from his long experience in Cultural Heritage (CH) and Humanities and "frustration" in the area of an emerging technical field, called Linked Data (LD) (watch a great 10+ min video for beginners by @manusporny). My frustration was just like Time-Berners Lee's story: he invented the World Wide Web, based on his frustration over data management (his vision of LD). LD is a simple set of design principles for the data on the web (or the data itself following the principles) which would take the web to a new level. It is not totally new any more. It has been increasingly popularized over the last decades, because it makes the web more database-oriented, so that we can ask questions and machines can give us answers more easily
That sounds like ChatGPT! That's true, but while ChatGPT is based on a Machine Learning with Large Language Model (LLM), LD is based on datasets mostly we humans designed and generated carefully. Although the invention like ChatGPT amazed us, one criticism is its black box approach. It is not easy to know how the machine gets the information. Biases caused by input data are also another question for Machine Learning. As it may give you convincing answers through dialogues with a user-friendly interface, there is a risk of over-trusting. On the other hand, LD could be more trustful in the sense we can know where the information comes from (if designed properly). In addition, the purposes are quite different. ChatGPT is not meant to integrate data on the web. It is currently rather a "standalone tool". LD has a formalized mechanism for reasoning to answer your questions. It is also noted that the two technologies may not be competing, but complementing each other. In fact, there is research to combine both
Technically speaking, LD allows us to create, publish, and share standardized structured data in a de-centralized manner to connect to each other through hyperlinks. It is a technology to move from human friendly "static" websites (documents) to a new form of web: "dynamic" machine-friendly interlinked graph databases of our knowledge, called Semantic Web. LD is designed to be its building block. While in Machine Learning, machines would "stealthily guess" answers to your question, by employing probability theories and a large amount of text processing, in Semantic Web, machines "logically and semantically interpret" data through inferences. In addition, LD is able to easily and smoothly integrate complicated and heterogeneous data scattered on the web. Consequently, we should be able to efficiently automate data processing on the web, which is not easy in many cases today. Great, right? However, there are still challenges to use LD. In the domain of CH and Humanities:
On top of these issues, the author feels the slow pace of LD adaptation in the web community in general. Although these are also related to more fundamental problems including LD quality and quantity, the author would like to address those challenges, provided we have a relatively desired LD environment, in order to demonstrate a scenario of and incentive to moving forward to our near-future
ReKisstory is my answer to tackle our challenges, although it is far from perfect. It is hoped to see a glimpse of the power of LD. This project also aims to attract more attention to LD within the web community to increase the level of benefits in the society
The technology is based on widely accepted and and cutting-edge standards
ReKisstory is the name chosen for the project. In the nutshell, it is a combination of Reki + Kiss + History. We would like to explore history with love. The ReKisstory app aims to show the way to step into time and space more easily and freely.
What is Reki? It is a Japanese character "歴", representing experience and record. Chinese basically has the same: "歷" or "历". History in Japanese is "Rekishi (歴史)". The logo of ReKisstory is based on the character "歴". Note ReKisstory also contains a word "history". We think it is also very important that users can build their own "stories" through "history".