We describe a method that detects the user's intentions on the basis of the user's operations on a digital map and automatically retrieves and displays appropriate Web pages. Since the series of operations that a user performs on digital maps usually reflect specific intentions, the user's intentions can be detected by analyzing the operation sequences. We describe basic operation sequences for detecting intentions and present examples of Web page retrieval using these sequences. We also explain how queries are generated based on the detected intentions and how the retrieved Web pages are displayed. Furthermore, we describe the implementation issues of a prototype system based on our method. In our evaluation experiments, we can retrieve the desired web pages by analyzing user's map operations. Finally we discuss the results of experiments focusing on the defined operation sequences, the detected intentions, and the retrieved and displayed Web pages.

No abstract

No abstract

No abstract

Presentation content can consist of heterogeneous media, such as videos and slides, which have been recorded and used. A variety of retrieval methods have been proposed for each kind of stored media. This paper proposes an extraction method using the features of each type of media. The proposed method specifically uses presentation video and slides. A user inputs query keywords, which are used to extract the most appropriate interval for the query according to appearance positions in the media. This enables the user to extract intervals that can be viewed more efficiently than when the interval depends only on the presence of query keywords.