Discussion Ontology: Knowledge Discovery from Human Activities in Meetings

PDF
Takahiro TSUCHIDA
Graduate School of Information Science, Nagoya University
Hironori TOMOBE
Graduate School of Information Science, Nagoya University
Shigeki OHIRA
EcoTopia Science Institute, Nagoya University
Katashi NAGAO
Center for Information Media Studies, Nagoya University

1 Introduction

In research activities, it is important to generate various ideas for the research theme through tasks such as It is possible that problems of bias of them to the intention of the individual, and the emanations by having a narrow view of things and the content, etc. occur though various ideas are created in the academic activity that is one of the activities of man's knowledge for the research theme through the work such as the document investigations and mounting. Then, the research theme is caught by a diversified aspect by presenting the idea to the research theme, and the discussion as others, the idea that emanates is made to be settled, and the policy of the activity in the future can be arranged.However, it is necessary to record the content of the discussion by some means to forget the content of the discussion with the passage of time.

It records in a semiautomatic target as the minutes of the multimedia that the content of the discussion activity like text information and the image and voice information, etc. is structurized, and authors research and are developing the technology that is called discussion Mining that extracts the knowledge reusable for man there. In this research, research activities is modeled as a cycle of two or more phases that surround the announcement and the discussion, and man in each phase and the relation to the minutes are analyzed.In addition, an interactive operation of the minutes according to each phase is examined, and it proposes the application that presses the activation of research activities.

2

The idea to the background and the problem of the research exists as a tacit concept of not expressible in a form and a logical language. The idea is presented to others to convert into the announced document that is the form from which those concepts are made a sign (Preparation phase), and to catch the research theme from a diversified aspect, and it discusses it while sharing (Discussion phase). Moreover, the content of the discussion is arranged, and the policy of the research is used to work like the investigation, mounting, and the verification, etc. ascertaining (Rumination phase) and afterwards (Investigation phase). And, the discussion is activated by reflecting a new idea and the problem that occurs from this work in the following announcement. These four phases think that research activities advances to be repeated gradually as shown in Figure 2. In this research, the model to whom the research activity is done by such a cycle is called DRIP model.

Fugure1:

There is SECI model as a relation research on this research. This is modeling of the process to which new knowledge is created by mutual action of "Form wisdom" whose transmission is possible knowledge by "Tacit knowledge" whose formalization and telling to others are difficult knowledge and a formal, logical languages. It is the one to achieve mutual of form wisdom and tacit knowledge action in the SECI model as an application at the minutes center in this research.

3

For the DRIP model, it is very important to do the research activity after the content of the discussion is based. However, it is a point that the index to forget the content of the discussion with the passage of time, and to refer to the discussion for becoming the problem here though the profitable discussion is remembered existing doesn't exist. What information in view of such a viewpoint, you give the minutes becomes important. Moreover, it is thought that those who inspect it give the minutes his interpretation when inspecting it, and it is possible to look back on the content of the discussion efficiently. And, the activation of the discussion can be expected to be pressed by taking the idea etc. invented from the work of mounting and the verification and consideration, etc. to the following announcement. Here, the application that supports the activation of research activities based on the minutes is described in each phase among DRIP models.

3.1

In discussion Mining that authors research and are developing, the discussion can be structurized by giving the meta data with various devices. Text information etc. that reflect the content of the information by segment and the remark of the image and the voice by the remark time are included in the meta data.As a result, recording the discussion structure composed of the remark linked with the image and the voice as the minutes becomes possible. Moreover, the index needed when efficient inspection is done by marking remark judged to be "It is important for me" can be given.

3.2

When the content of the discussion is ruminated, the image and the voice linked with the mark remark can be inspected. However, information tedious as repeated an interjection, silent time, and the same content is included there. Those who inspect it give the summary of the content of the remark as meta information (annotation) to extract information that is necessary from these information and it is possible to do. This annotation can be given to not only one remark but also the settlement of the remark. As a result, the policy of the activity until discussing it as follows is expected to be able to be understood.

3.3

It is very important to do the research activity after the discussion that has been done in the past is based. In a word, the annotation made by the Rumination phase should be able to be referred to easily.Therefore, a usual Web browser is used in the application proposes by this research and the list of the annotation can be referred to. Moreover, a more detailed content of the discussion can be confirmed by inspecting the image and the voice linked from each annotation with the discussion about the object.

A new discovery, the idea, and the problem that occurs while mounting, verifying, and evaluating it while considering the content of the discussion thus have the key role as a topic when the following discussion is done. However, it forgets with the passage of time as long as these neither idea nor problem are preserved by some forms. Therefore, the idea and the problem can be recorded by the diary form by using the interface like Figure 2 in this research.

Fugure2:

3.4

It was accumulated by the Investigation phase, and it is thought that taking in material that uses the idea and the problem that occurs newly from the document investigation and a current mounting and the verification work for the following announcement without lack is effective to enhance the content of the discussion. The accumulated list can be converted into the Microsoft PowerPoint document for that.

4 Conclusions

In this research, it proposed the application to press the activation of research activities by modeling research activities as a cycle of four phases, and analyzing the relation to the minutes.The following points are enumerated as a problem in the future.