Yukio Ohsawa

KeyGraph: automatic indexing by co-occurrence graph based on building construction metaphor

Presents an algorithm for extracting keywords representing the asserted main point in a document, without relying on external devices such as natural-language processing tools or a document corpus. Our algorithm, KeyGraph, is based on the segmentation of a graph, representing the co-occurrence between terms in a document, into clusters. Each cluster corresponds to a concept on which an authors idea is based, and the top-ranked terms are selected as keywords using a statistic based on each terms relationship to these clusters. This strategy comes from considering that a document is constructed like a building for expressing new ideas based on traditional concepts. The experimental results show that the thus-extracted terms match the authors main point quite accurately, even though KeyGraph does not use each terms average frequency in a corpus, i.e. KeyGraph is a content-sensitive, domain-independent indexing device.

New Frontiers in Artificial Intelligence

Neg-Raising (NR) verbs form a class of verbs with a clausal complement that show the following behavior: when a negation syntactically attaches to the matrix predicate, it can semantically attach to the embedded predicate. This paper presents an account of NR predicates within Tree Adjoining Grammar (TAG). We propose a lexical semantic interpretation that heavily relies on a Montague-like semantics for TAG and on higher-order types.

Chance discoveries for making decisions in complex real world

Chance discovery is to become aware of a chance and to explain its significance, especially if the chance is rare and its significance is unnoticed. This direction matches with various real requirements in human life. This paper presents the significance, viewpoints, theories, methods, and future work of chance discovery. Three keys for the progress are extracted from fundamental discussions on how to realize chance discovery: (1) communication, (2) imagination, and (3) data mining. As an approach to chance discovery, visualized data mining methods are formalized as tools aiding chance discoveries on the basis of these keys.

DATA CRYSTALLIZATION: CHANCE DISCOVERY EXTENDED FOR DEALING WITH UNOBSERVABLE EVENTS

This paper introduces the concept of chance discovery, i.e. discovery of an event significant for decision making. Then, this paper also presents a current research project on data crystallization, which is an extension of chance discovery. The need for data crystallization is that only the observable part of the real world can be stored in data. For such scattered, i.e. incomplete and ill-structured data, data crystallizing aims at presenting the hidden structure among events including unobservable ones. This is realized with a tool which inserts dummy items, corresponding to unobservable but significant events, to the given data on past events. The existence of these unobservable events and their relations with other events are visualized with KeyGraph, showing events by nodes and their relations by links, on the data with inserted dummy items. This visualization is iterated with gradually increasing the number of links in the graph. This process is similar to the crystallization of snow with gradual decrease in the air temperature. For tuning the granularity level of structure to be visualized, this tool is integrated with humans process of chance discovery. This basic method is expected to be applicable for various real world domains where chance-discovery methods have been applied.

KeyGraph: Visualized Structure Among Event Clusters

The most fundamental causes may be hidden and in severe cases unknown (not in the knowledge of a human nor a computer). These causal events might be occurring eternally, or be brought up from a sequence in the past and trigger events in the future. Here is presented KeyGraph, generalized from a document-indexing method to a method for extracting essential events and the causal structures among them from an event sequence.

Modeling the Process of Chance Discovery

The fundamental philosophy of chance discovery is introduced. By comparison with the cyclic model of knowledge discovery, this chapter describes the essentials for realizing chance discovery. From these discussions, three keys for chance discovery are proposed, i.e. communication, context shifting, and data mining. As a result, the double helix and the subsumption architecture are presented as methods for realizing chance discovery.

Chance Discovery By Stimulated Groups Of People. Application To Understanding Consumption Of Rare Food

Chance discovery is to become aware of and to explain the significance of a, that is, a piece of information about events or situations that is significant for decision making. Sometimes a chance is rare and its significance is unnoticed. This paper proposes a method to merge three keys for chance discovery: (1) communication; (2) imagination and (3) data mining. Applied to the case of meal service, a visualised data mining method is used for discovering unnoticed demands underlying family consumption behaviour. The visualised relations between usual and unusual consumption patterns stimulate the awareness of and the communication among housewives talking in a room. This leads to the discovery of latent family consumption demands and to the proposal of serving meals that have unnoticed significant merits for their families.

Data Jackets for Synthesizing Values in the Market of Data

Abstract In order to make a social environment where analysts and decision makers in active businesses and sciences can be provided with data they need, we discuss why we should and how we can (re)design an environment called the market of data, where each user or provider of data can externalize and share the value of each part of data so that one can buy/sell it in a reasonable condition, e.g., for a reasonable price or as open source if it may give merits to people in general rather than to particular segments who can pay. Presenting one way to go, that is called Innovators’ Marketplace on Data Jackets, we show a vision for the systematic design of the market of data where existing tools and new technologies are configured, such as the visualization of the relation among databases/datasets for aiding stakeholders’ communication about possible use scenarios of data. By this, this paper calls for potential contributors in the future to the market of data.

Idea discovery: A scenario-based systematic approach for decision making in market innovation

A new trend of researches on knowledge discovery and chance discovery is to identify human insights through data synthesis rather than to discover facts through data analysis. In this paper, we propose a systematic approach named idea discovery which is committed to turning data into effective human insights. Idea discovery focuses on dynamic and sustainable process for high-quality ideas cultivation, construction, integration and evaluation through human-computer and human-human interaction. It mainly relies on latent information and its dynamic changes to drive ideas creation, integration and evaluation during sustainable creativity process. The process of idea discovery is in accordance with a dynamic model which contains two key components: (1) mining algorithms to turn data into scenario maps for eliciting human insights; (2) scenario-based creativity support activities towards actionable ideas generation. An intelligence system called Galaxy integrated with IdeaGraph algorithm has been developed to support the dynamic process of idea discovery. A case study in an automobile company has validated the effectiveness of proposed method and system.

Innovation support system for creative product design based on chance discovery

Under a turbulently changing and highly competitive market, discovery of a chance is always significant for many companies to launch new and creative products or services in time, fulfilling consumers demands for occupying more market share. Many available methods on market research for designing new products are more focused on the analysis process, so that product designers run short of ideas discovery. In this paper, we present a novel innovation support system (ISS) based on chance discovery with data crystallization to assist human innovation in designing new products, especially creative products. The ISS is a human-centric system to enable value cognition and follows the following process: (1) visualized scenario graph generation, (2) human value cognition, (3) value co-creation based on shared knowledge, and (4) emerging chances evaluation. The result of a case study validates the effectiveness of ISS.