JPS6370297A - Information processor - Google Patents

Abstract

(57) [Abstract] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] [Object of the invention] (Industrial application field) The present invention provides an information processing device I that can be associated with an object.
Regarding this.

(Prior Art) With the recent development of speech recognition technology, it is becoming possible to recognize vocal sounds fairly accurately.

As a result, tags using voice recognition technology have come to be used for a variety of purposes. For example, a simple key chain has a device (tag) that responds when you clap your hand. This system recognizes the sound of clapping hands and generates an electronic sound, allowing you to know where the keychain is. (Fig. 8) However, in this case, since it only responds to a special signal, the sound of clapping hands, it also responds to other signals, so it registers all the signals it should respond to in the dictionary section and reacts. All signals must be estimated with pointers that represent what should be done. However, this has the limitation that it can only respond to special signals and names that are registered and presumed to be pointers. With the recent development of AI research, these problems can be solved by positioning the object in a hierarchical relationship with the object itself and other conceptually related objects.

(Problems to be Solved by the Invention) As described above, conventional tags have the disadvantage that they can only respond to registered unique signals and names.

The present invention has been made in consideration of these circumstances, and its purpose is to use not only pre-registered words to express stored items but also more general words by associating them with stored items. An object of the present invention is to provide an information processing device (tag) that can respond to words and tell people where they are stored.

[Structure of the invention]

(Means for Solving the Problems) The present invention includes a speech recognition section that recognizes speech, a dictionary section that stores a predetermined word in association with a word that is a superordinate concept of this word, and a speech recognition section that recognizes speech. The present invention is characterized by comprising a reaction section that reacts when the word recognized by the recognition section matches a predetermined word stored in the dictionary section or a word in a superordinate concept of this word. It is something.

(Function) The present invention has the words expressing the objects to be associated in a hierarchy of concepts related to the objects in a pre-registered dictionary, so that the words that are uttered are limited. People will respond to words that are connected not just by words but also by relationships.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic diagram of an embodiment of the system. 1 is a speech recognition unit that recognizes the uttered voice; 2 is a dictionary unit in which the names of objects and their hierarchical relationships are registered in advance; 3 is 1
This is a reaction part that reacts only when the word recognized in the dictionary satisfies the relationship with the word registered in the dictionary. In addition, 2 is a schematic configuration diagram of the dictionary section, in which 4 is an extraction section that cuts out a series of words recognized by the recognition section 1 word by word, and 5 is the name of the object that has been associated with it in advance. This is a storage unit that stores the hierarchical relationship including.

Now, the information regarding the object stored in advance in the storage unit 5 is a tree structure representing the relationship between the object and its superordinate concept, concrete object, and the like. As an example, assume that the object is a "floral patterned plate."

In other words, let's assume that this device is stored together with a ``floral plate''. First of all, ``Floral pattern plate J'' is a combination of two words: ``floral pattern'' and ``plate J.''Furthermore, ``floral pattern'' has the property of ``pattern,'' and ``plate'' has the property of ``Western tableware.'' It has the property of J.

Furthermore, ``Western tableware J has the property of ``tableware J'' (Figure 2).If such a relationship is expressed in a tree structure, it will be as shown in Figure 3, but the storage section has created such a tree structure in advance. It is held as information.Also, in this tree structure, a pointer l to the object
(Figure 4). Pointer 1 indicates whether this device is associated with the device.

This will be explained below using FIG. 5.

First, if the voice recognized in Figure 1
Then, in the cutout part 4, from this word "floral pattern" and "
Cut out the word plate J (steps a, b). In the storage section 5, it is searched to see if these "floral pattern J" and "plate j" exist in the tree structure registered in advance (step C), and if they exist, they are pointed to with the corresponding number of pointers (step d). ).

In this case, as shown in Figure 6, the pointers point to "floral pattern" and "plate j."In this case, pointer 1, which represents the associated object, points to "floral pattern" and "plate j." pointing. Since the pointer pointing to "floral pattern" and the pointer pointing to "dish" exist at the same position as pointer 1 (step C), a signal is sent to the reaction section (step f) to perform a reaction. In other words, the device stored together with the object ``Flower-patterned plate J'' responded to the voice ``Floral-patterned plate J'' and indicated the storage location.

Even if the word ``patterned Western tableware J,'' which is a superordinate concept of ``floral patterned plate J,'' is recognized in the wooden structure in the storage section of 5, first, in 4, the words ``patterned J'' and ``Western food BJ'' are recognized. Cut it out and point it with the pointer in the same way as in the above example.
Since pattern J and "Western tableware" are located, a reaction is performed. In other words, it is possible to react even to words expressing superordinate concepts.

As another example, for ``red western tableware with a floral pattern'', it is cut out ``floral pattern J'', ``red j'', and among the three words of Western tableware j, the word ``red J'' does not exist in the tree structure, so it is an object object. I conclude that the word ``is not a word that describes a flower-patterned plate j'' and end the sentence.

As mentioned above, if you register the relationship between concepts representing objects in a dictionary in a tree structure, and have only pointers representing objects and pointers representing recognized words, the tree structure is followed. This makes it possible to provide an intelligent device that responds to any combination of words representing objects.

Furthermore, in this embodiment, by allowing the user to freely define the tree structure in the storage section, it is possible to make the device personal to the user. Furthermore, by providing a plurality of tree structures in the storage section and by providing a plurality of pointers 1 representing objects in the tree structure, it is possible to associate one device with a plurality of objects.

〔Effect of the invention〕

Thus, according to the present invention, when searching for a stored item, the system not only responds when you call out the specific name of the item and tells you where it is stored, but also responds to the item's more general name or name representing a superordinate concept. This makes it possible to perform intelligent processing.

[Brief explanation of the drawing]

FIG. 1 is a schematic configuration diagram of a device according to an embodiment of the present invention, FIG. 2 is a table showing the conceptual structure of words registered in advance in the storage section of the device according to the embodiment, and FIG. FIG. 4 is a diagram showing the conceptual structure in this structure; FIG. 4 is a diagram including the pointer 1 representing the object in FIG. 3; FIG. 5 is a diagram showing the processing flow of the dictionary section in the embodiment device; FIG. is a diagram showing a new pointer generated when the example in FIG. 4 is implemented, FIG. 7 is a diagram showing a case where an example different from FIG. 6 is implemented, and FIG. FIG. 1... Speech recognition section, 2... Dictionary section, 3... Reaction section, 4... Word extraction section, 5... Storage section.

Claims (2)

[Claims] (1) A speech recognition section that recognizes speech; a dictionary section that stores a predetermined word in association with a word that is a superordinate concept of the word; and a word recognized by the speech recognition section.
An information processing device comprising: a reaction section that reacts when a predetermined word stored in the dictionary section or a word in a superordinate concept of this word matches. (2) The information set forth in claim 1, in which the speech recognition section, dictionary section, and reaction section are integrated and housed together with objects that correspond to predetermined words in the dictionary. Processing equipment.