"Where are the Christmas Decorations?": A Memory Assistant for Storage Locations

Michael Van Hilst; Lewis G. Creary

"Where are the Christmas Decorations?": A Memory Assistant for Storage Locations

Michael Van Hilst
, Lewis G. Creary

Research output: Contribution to conference › Presentation

Abstract

At Hewlett-Packard Laboratories we want to know how inexpensive it can be to endow mobile personal assistants with the ability to speak naturally with their users. To this end, we are investigating and demonstrating speech-capable mobile personal assistants that can be realized using common off-the-shelf hardware and software components, a modest development effort, and wireless connectivity. As a part of this activity, we built a storage location memory assistant by connecting a voice front end on a PDA to a database back end on a remote server, with speech, natural language, and dialog processing in between. Specifically, the storage location memory assistant saves and retrieves information about the locations of stored objects in and around the user's house. Here is a sample dialog:

User: Where are the Christmas decorations?

PDA: They're in the leftmost medium-sized white box under the wood table in the garage.

To implement the storage location memory assistant, we leverage technologies in speech recording, voice recognition, parser generation, and database management to produce domain-limited natural language understanding and semi-structured knowledge representation. In this paper, we describe the architecture of the system, and some of the more interesting technical details.

Original language	American English
State	Published - Jun 1 2001
Event	6th International Workshop on the Applications of Natural Language to Information Systems - Duration: Jun 1 2001 → …

Conference

Conference	6th International Workshop on the Applications of Natural Language to Information Systems
Period	6/1/01 → …

Disciplines

Computer Sciences

Cite this

@conference{2d8b810a7c3646468db3217a6aa8a299,

title = "{"}Where are the Christmas Decorations?{"}: A Memory Assistant for Storage Locations",

abstract = " At Hewlett-Packard Laboratories we want to know how inexpensive it can be to endow mobile personal assistants with the ability to speak naturally with their users. To this end, we are investigating and demonstrating speech-capable mobile personal assistants that can be realized using common off-the-shelf hardware and software components, a modest development effort, and wireless connectivity. As a part of this activity, we built a storage location memory assistant by connecting a voice front end on a PDA to a database back end on a remote server, with speech, natural language, and dialog processing in between. Specifically, the storage location memory assistant saves and retrieves information about the locations of stored objects in and around the user's house. Here is a sample dialog: User: Where are the Christmas decorations? PDA: They're in the leftmost medium-sized white box under the wood table in the garage. To implement the storage location memory assistant, we leverage technologies in speech recording, voice recognition, parser generation, and database management to produce domain-limited natural language understanding and semi-structured knowledge representation. In this paper, we describe the architecture of the system, and some of the more interesting technical details.",

author = "\{Van Hilst\}, Michael and Creary, \{Lewis G.\}",

year = "2001",

month = jun,

day = "1",

language = "American English",

note = "6th International Workshop on the Applications of Natural Language to Information Systems ; Conference date: 01-06-2001",

}

TY - CONF

T1 - "Where are the Christmas Decorations?": A Memory Assistant for Storage Locations

AU - Van Hilst, Michael

AU - Creary, Lewis G.

PY - 2001/6/1

Y1 - 2001/6/1

N2 - At Hewlett-Packard Laboratories we want to know how inexpensive it can be to endow mobile personal assistants with the ability to speak naturally with their users. To this end, we are investigating and demonstrating speech-capable mobile personal assistants that can be realized using common off-the-shelf hardware and software components, a modest development effort, and wireless connectivity. As a part of this activity, we built a storage location memory assistant by connecting a voice front end on a PDA to a database back end on a remote server, with speech, natural language, and dialog processing in between. Specifically, the storage location memory assistant saves and retrieves information about the locations of stored objects in and around the user's house. Here is a sample dialog: User: Where are the Christmas decorations? PDA: They're in the leftmost medium-sized white box under the wood table in the garage. To implement the storage location memory assistant, we leverage technologies in speech recording, voice recognition, parser generation, and database management to produce domain-limited natural language understanding and semi-structured knowledge representation. In this paper, we describe the architecture of the system, and some of the more interesting technical details.

AB - At Hewlett-Packard Laboratories we want to know how inexpensive it can be to endow mobile personal assistants with the ability to speak naturally with their users. To this end, we are investigating and demonstrating speech-capable mobile personal assistants that can be realized using common off-the-shelf hardware and software components, a modest development effort, and wireless connectivity. As a part of this activity, we built a storage location memory assistant by connecting a voice front end on a PDA to a database back end on a remote server, with speech, natural language, and dialog processing in between. Specifically, the storage location memory assistant saves and retrieves information about the locations of stored objects in and around the user's house. Here is a sample dialog: User: Where are the Christmas decorations? PDA: They're in the leftmost medium-sized white box under the wood table in the garage. To implement the storage location memory assistant, we leverage technologies in speech recording, voice recognition, parser generation, and database management to produce domain-limited natural language understanding and semi-structured knowledge representation. In this paper, we describe the architecture of the system, and some of the more interesting technical details.

UR - https://nsuworks.nova.edu/gscis_facpres/568

M3 - Presentation

T2 - 6th International Workshop on the Applications of Natural Language to Information Systems

Y2 - 1 June 2001

ER -

"Where are the Christmas Decorations?": A Memory Assistant for Storage Locations

Abstract

Conference

Disciplines

Other files and links

Fingerprint

Cite this