JP5129999B2 - Objective estimation device, objective estimation method and program - Google Patents

Description

  The present invention provides an object estimation method for supporting user operations in home appliances such as a computer having a menu selection type operation interface, a DVD (Digital Versatile Disk) recorder, a digital television, a mobile phone, and a car navigation system. More particularly, the present invention relates to a purpose estimation device and a purpose estimation method for estimating a user's purpose from a selection operation and providing it as support information.

  In recent years, the operation of electrical appliances such as computers, mobile phones, and DVD recorders has become complicated with the implementation of various functions. Therefore, by adopting a selection-type menu-type interface, particularly a hierarchical interface, in the device (electric appliance device), it is attempted to suppress the increase in complexity and make it easy to use many functions.

  However, in such a hierarchical interface, the choice words displayed on the menu are difficult for the user to understand, and the concept of the words displayed on the choices deviates between the interface designer and the user who uses the device. There is. For example, in the operation of a DVD recorder, the option that the user thinks of performing the task “I want to make subtitles English” and the option that is actually displayed deviate, and the user cannot find the desired function. The problem is surfaced. Therefore, effective users are able to overcome the confusion of not being able to find a function caused by the difference between the concept of the designer's task and the concept of the user's task inherent in such a selection type menu-type interface. It is hoped that support will be realized.

  Conventionally, an apparatus for solving such a problem has been proposed (see, for example, Patent Document 1). The apparatus of Patent Document 1 assigns a keyword related to each function prepared in the device, and when a user inputs a keyword related to a target function by pressing the help key, the function corresponding to the keyword Search and present (Help search). Therefore, in the apparatus of Patent Document 1, it is necessary to register in advance all keywords that the user may input in association with functions. However, it is difficult to anticipate and assign all the keywords as described above, and there are cases where the user cannot input an appropriate keyword. As a result, the apparatus disclosed in Patent Document 1 cannot properly estimate the function intended by the user.

  Therefore, in order to solve the above-described problems, a purpose estimation device that estimates the user's purpose from the operation history of the device by the user instead of the user inputting a keyword has been proposed (for example, Patent Documents). 2).

  FIG. 31 is a block diagram showing the configuration of the purpose estimation apparatus of Patent Document 2.

  This purpose estimation apparatus includes an input unit 3201, a request statement storage unit 3202, a search unit 3203, a display unit 3204, a history information storage unit 3205, an operation history vector generation unit 3206, and a function explanation database 3207. ing.

  The input unit 3201 includes a keyboard, a mouse, and the like, and accepts operation information and keywords for help search. The history information storage unit 3205 stores operation information received by the input unit 3201. The operation history vector generation unit 3206 generates a feature vector corresponding to the operation history using the operation information of the previous n steps stored in the history information storage unit 3205 and the operation vector dictionary.

The search unit 3203 calculates an inner product between the feature vector corresponding to the operation history and the function explanation vector assigned to each function stored in the function explanation database 3207, and each function for the operation history (feature vector). Find the score. Further, the search unit 3203 also calculates a score by help search, and searches for a function to be helped, that is, a function intended by the user, by combining both scores. The display unit 3204 displays the functions as search results of the search unit 3203 in the order of score.
Japanese Patent Laid-Open No. 7-121563 JP 2000-137726 A

  However, the purpose estimation device disclosed in Patent Document 2 has a problem in that it cannot properly estimate the user's purpose.

  In other words, this purpose estimation device calculates the inner product of the feature vector that uniquely indicates the operation history and the function explanation vector that uniquely indicates the function for each function, so that the operation history (operation information) and each function are calculated. Compare similarities of. Therefore, this purpose estimation apparatus compares the operation information that is not originally specified from the operation information that is the input option among all the elements included in the function. For example, the operation information indicates “program guide”, and the functions include “display program guide” and “reserve recording from program guide”. In this case, the operation information “program guide” alone does not indicate whether the user wants to view the program guide or wants to make a recording reservation from the program guide. Nevertheless, the purpose estimation apparatus of Patent Document 2 uses the functions “display program guide” and “reserve recording from the program guide” as unique function explanation vectors, respectively, and only the operation information “program guide”. The comparison including the elements “display” and “recording reservation” which cannot be specified originally from the user is performed. As a result, in this purpose estimation device, the superiority or inferiority between the functions is more than necessary, and the function intended by the user cannot be estimated appropriately.

  Hereinafter, such a problem will be described in detail.

  In general, in a display-based human computer interface (HCI), a user evaluates information displayed on a display from the viewpoint of a task goal, performs a selection operation, and observes and evaluates a change resulting from the operation. Perform the next selection operation. At this time, the hierarchical interface has a feature that the function is specified by going down the hierarchy, and therefore the range of functions that can be specified from a single input is limited. For example, if the subject wants to set the time of the clock in front of the DVD recorder, the user can go down the hierarchy like “Function selection → Other functions → Initial setting → Clock setting → Time adjustment”. The functions are gradually specified as “setting” → “clock” → “time”. Therefore, the range that can be specified from one operation is limited.

  On the other hand, when the user (subject) is given a task (desired function) such as “Erase the recorded program” and the behavior of the subject is observed, the “Browsing the program guide” option is selected. “Browsing”, which is different from “,” was frequently observed, and the behavior of selecting only the word “program” was frequently observed. That is, it is considered that the subject is performing a selection operation using a part of the words included in the options as a clue.

  Therefore, if the user selects an option described only as “program guide”, the range that can be specified is limited as described above, so whether the user wants to watch the program or wants to make a recording reservation. However, it is not known whether the user wants to confirm the reception schedule in the program guide, and only knows that he wants to operate the program or the program guide. In other words, even if there are functions such as “display program guide”, “reserve recording from program guide”, and “check reception schedule of program guide”, only the information of “program guide” selected by the user is used. , You can't make these features superior or inferior.

  Nevertheless, the purpose estimation device of Patent Document 2 always makes comparisons between the overall characteristics of the functions and the options (operation information or operation history). I will put it on.

  Therefore, the present invention has been made in view of such a problem, and provides a purpose estimation device capable of appropriately estimating a function intended by a user among a plurality of functions of an operation target device. With the goal.

In order to achieve the above object, the purpose estimating apparatus according to the present invention includes, for each function of the operation target device, option acquisition means for acquiring an option selected by a user from among a plurality of options as an input option. A function element storage unit that stores a plurality of elements included in the character string of the function name of the function as an element group, and one or more of the function element groups for each function of the operation target device A subgoal that specifies a subrange corresponding to the maximum similarity as a subgoal by calculating a similarity between each of the plurality of partial ranges composed of the elements and the input option acquired by the option acquisition unit A specifying unit, and an estimation unit that estimates that a function having a higher degree of similarity of the subgoal among a plurality of functions of the operation target device is intended for the user. And features.
In addition, in order to achieve the above object, the purpose estimation device according to the present invention includes a plurality of functions included in an operation target device that executes a function corresponding to the plurality of options when a plurality of options are selected. , A purpose estimation device for estimating a user's target function, an option acquisition means for acquiring an option selected by the user from among a plurality of options as an input option, and for each function of the operation target device, A function element storage unit that stores a plurality of elements included in the character string of the function name of the function as an element group, and one or a plurality of element groups of the function for each function of the operation target device A plurality of partial ranges different from each other, and calculating the similarity between the input option acquired by the option acquisition means and each of the plurality of partial ranges, Subgoal specifying means that enables specifying a range that can be estimated from input options by specifying a subrange corresponding to the similarity as a subgoal, and among the functions of the operation target device, the similarity of the subgoal It is characterized by comprising estimation means for estimating that a function having a larger value is intended for the user.

  As a result, when an option is selected by the user, for each function of the operation target device, a partial range most relevant to the option (input option) is specified as a subgoal from among the function names of the function. It is estimated that a function having a higher degree of similarity is a function intended by the user. When a subgoal is specified, for example, by using an element (partial element) such as a word included in the function name character string, from one function name to a partial range (subgoal candidate) consisting of one word, Subranges consisting of words (subgoal candidates) are set, and among those subranges, the subrange most closely related to the input option selected by the user, that is, the subrange having the highest similarity between the input options Is identified as a subgoal. In other words, it is estimated which partial range of each function name the user selected the input option.

  Therefore, for example, when there is a function name “display program guide” and a function name “reserve recording from the program guide”, when the user selects “program guide” as an input option, the function name “program guide For “Display”, the partial range “Program guide” and “Display” are set, and for the function name “Reserve recording from the program guide”, the partial range “Program guide” and “Recording reservation” are set. Is set.

  Then, among the partial ranges “program guide” and “display” of the function name “display program guide”, the partial range “program guide” having the highest similarity to the input option “program guide” is specified as the subgoal. The In other words, if the user is aiming at the function “display program guide”, it is presumed that the user selects the input option “program guide” by paying attention to the function name “program guide”. Similarly, among the partial ranges “program guide” and “recording reservation” of the function name “program recording reservation”, the partial range “program guide” having the highest similarity to the input option “program guide” is Identified as a subgoal. In other words, if the user is aiming at the function “reserve recording from the program guide”, it is presumed that the user selects the input option “program guide” by paying attention to the “program guide” of the function name.

  In this case, the similarity of the sub-goal “program guide” of the function name “display program guide” for the input option “program guide” and the sub-goal of the function name “program recording reservation from the program guide” for the input option “program guide” The similarities of the “program guide” are calculated to be equal to each other.

  When a subgoal is specified in this way, it is estimated that a function having a higher degree of subgoal similarity between the function “display program guide” and the function “reserve recording from the program guide” is the user's purpose. Is done. Therefore, the function “display program guide” and the function “reserve recording from the program guide” are estimated to be the user's purposes to the same extent.

  On the other hand, conventionally, the function “display the program guide” and the function “reserve recording from the program guide” are treated as one function explanation vector different from each other, and the function explanation vector and the feature vector of the operation history Are compared with each other, and more than necessary superiority or inferiority is assigned between the function “display program guide” and the function “reserve recording from the program guide”. In other words, the function name “display program guide” is used even though the user does not know whether the user wants to view the program guide or schedule recording from the program guide from the input option “program guide” selected by the user. Comparison is made with the operation history including “display” of “Yes” and “recording reservation” of the function name “reserve recording from the program guide”. As a result, one of the function “display program guide” and the function “reserve recording from the program guide” is estimated as the user's purpose.

  However, in the present invention, as described above, the functions intended by the user can be appropriately estimated without imparting more than necessary superiority or inferiority to each function.

  In addition, the subgoal specifying unit generates the plurality of elements as input interpretations by dividing the character string expressing the input option acquired by the option acquisition unit into a plurality of elements, and the function of the operation target device Each calculating a similarity between each of the plurality of input interpretations of the input options and each of the plurality of partial ranges, and specifying the partial range corresponding to the maximum similarity as the subgoal. It is good.

  For example, when the user selects an option, the option is selected by paying attention to any element in a character string expressing the option. Therefore, there are cases where an extra element ignored by the user is included in the character string of the option selected by the user.

  Therefore, in the present invention, the similarity is calculated between each of the plurality of input interpretations of the options (input options) and each of the plurality of partial ranges of the function, and the partial range corresponding to the maximum similarity is the subgoal. Therefore, by omitting an extra element in the input options, it is possible to appropriately estimate the function intended by the user without making the function more than necessary.

  In addition, the subgoal specifying unit specifies, for each input option acquired by the option acquisition unit, a subgoal of each function of the operation target device for the input option, and the estimation unit is configured for each function of the operation target device. In addition, the maximum similarity among the plurality of subgoal similarities specified between the function and each input option is calculated as the possibility score of the function, and the plurality of functions of the operation target device are calculated. Among them, it is possible to provide a possibility-based estimation unit that estimates that a function having a higher possibility score is intended for the user.

  For example, when the choices “program guide” and “recording reservation” are selected as input choices, the subgoal “program guide” and its input between the function “reserve recording from the program guide” and the input choice “program guide” The subgoal similarity is specified, and the subgoal “recording reservation” and the similarity of the subgoal are specified between the function “reserve recording from the program guide” and the input option “recording reservation”. In this case, the maximum similarity between the similarity of the subgoal “program guide” and the similarity of the subgoal “recording reservation” is the function “program guide” for the function “reserve recording from the program guide”. Is calculated as a possibility score of “Reserve recording from”. Such a possibility score is calculated for all functions of the operation target device, and a function having a larger possibility score is estimated to be the user's purpose.

  Therefore, even if the number of input choices selected by the user is small, it is possible to appropriately estimate the function intended by the user.

  In addition, the subgoal specifying unit specifies, for each input option acquired by the option acquisition unit, a subgoal of each function of the operation target device for the input option, and the estimation unit is configured for each function of the operation target device. In addition, the larger the ratio of the number of elements not included in any of the plurality of subgoals specified between the function and each input option in the element group of the function, the smaller the plurality of subgoals. Belief-based estimation means for calculating a certainty factor that increases as the degree of similarity increases, and estimates that the function having the greater certainty factor among the functions of the operation target device is intended for the user. It is good also as providing.

  For example, when the options “program guide” and “recording reservation” are selected as input options, the subgoal “program guide” and the subgoal of the subgoal between the function “display program guide” and the input option “program guide” are selected. The similarity is specified, and the subgoal “program” and the similarity of the subgoal are specified between the function “display program guide” and the input option “recording reservation”.

  In this case, the function “display program guide” is not included in either the subgoal “program guide” or the subgoal “program” in the element group of the function name “display program guide”. A certainty factor is calculated such that the larger the ratio of the number of elements (for example, “display”) is, the smaller the degree is, and the larger the degree of similarity of the subgoal “program guide” and the degree of similarity of the subgoal “program” is.

  At this time, since the similarity of the subgoal “program” is small and the ratio of the number of elements not included in any of the subgoals is large, the certainty factor is calculated to be relatively small.

  Such a certainty factor is similarly calculated for other functions.

  That is, when the choices “program guide” and “recording reservation” are selected as input choices, the subgoal “program guide” and its sub-goals between the function “reserve recording from the program guide” and the input choice “program guide” The subgoal similarity is specified, and the subgoal “recording reservation” and the similarity of the subgoal are specified between the function “reserve recording from the program guide” and the input option “recording reservation”.

  In this case, for the function “Reserve recording from the program guide”, the subgoal “Program guide” and the subgoal “Record reservation” are selected from the element group of the function name “Reserve recording from the program guide”. The certainty factor is calculated such that the smaller the ratio of the number of elements not included, the smaller the ratio, and the greater the similarity of the subgoal “program guide” and the similarity of the subgoal “recording reservation”.

  At this time, since the similarity of each subgoal is large and the ratio of the number of elements not included in any of the subgoals is small, the certainty factor is calculated to be relatively large.

  That is, in the case of the function “reserve recording from the program guide”, a greater certainty factor is calculated than the function “display program guide” described above.

  As a result, when the choices “program guide” and “recording reservation” are selected as input choices, the function “reserve recording from the program guide” is more user-friendly than the function “display program guide”. It is estimated that

  As described above, in the present invention, the certainty factor increases as the subgoal is specified in a wider range of the function name character string, and thus the function intended by the user can be estimated more appropriately and accurately.

  The purpose estimation apparatus further includes support request receiving means for receiving a support request from a user, and the subgoal specifying means is the input acquired by the option acquiring means until the support request is received. For each option, a subgoal of each function of the operation target device with respect to the input option is specified, and the possibility-based estimation unit or the confidence-based estimation unit of the estimation unit is configured to accept the request for assistance. The possibility score or the certainty factor may be calculated, and a function having a higher possibility score or certainty factor may be estimated to be a user's purpose.

  Thereby, when a request for assistance is received from the user, that is, when the user is confused about selection of an option, the function estimated as the user's purpose can be presented at an appropriate timing.

  The purpose estimation apparatus further includes a presentation unit that presents the function estimated by the estimation unit, and the estimation unit further includes, for each function of the operation target device, the function and the input options. Among the plurality of subgoals similar to each other, the maximum similarity is calculated as the possibility score of the function, and the function having the highest possibility score is selected from the plurality of functions of the operation target device. The possibility-based estimation unit that estimates that the user's purpose is assumed, and when the certainty factor is larger than the first threshold value, causes the presenting unit to present the function estimated by the certainty factor-based estimation unit. Switching means for causing the presenting means to present the function estimated by the possibility-based estimating means when the certainty factor is smaller than the first threshold value. It may be.

  For example, when the number of input options selected by the user is small, the certainty factor is smaller than the first threshold, and when the number of input options selected by the user is large, the certainty factor is larger than the first threshold value. There is a tendency. Therefore, when the certainty factor is smaller than the first threshold value, a function with a higher possibility score is estimated as the user's purpose based on a small number of clues, and obtained when the certainty factor is larger than the first threshold value. It is estimated that the user's purpose is a function having a higher certainty factor by effectively using the clue. As a result, the function intended by the user can be appropriately estimated according to the status of the input option selected by the user.

  In addition, the certainty factor base estimation unit updates the certainty factor of each function each time input options are sequentially acquired by the option acquisition unit, and the purpose estimation device further includes a function estimated by the estimation unit. When the overall change amount of the certainty factor of each function updated by the certainty factor-based estimating unit is smaller than a second threshold, the estimation factor is estimated by the certainty factor-based estimating unit It is good also as providing the presentation timing determination means which makes the said presentation means show a function.

  For example, if the user is stuck in selecting an option, the amount of change in confidence decreases. Therefore, in the present invention, when the overall change amount of the certainty factor of each function, for example, the change amount of the average value of the certainty factor of each function is smaller than the second threshold, the function intended by the user is presented. Therefore, when the user is stuck in selecting an option, the function estimated as the user's purpose can be presented at an appropriate timing without accepting a request for assistance from the user. As a result, it is possible to suppress the act of the user repeatedly selecting choices while feeling stuck in the choice of choices.

  In addition, the presentation timing determination unit may be configured to use the certainty factor base when the change amount is smaller than the second threshold value and there is a certainty factor larger than a third threshold value among the certainty factors of the respective functions. The function estimated by the estimating means may be presented to the presenting means.

  As a result, when any certainty factor is larger than the third threshold, that is, when the user's target function is estimated sufficiently correctly, the function is presented, so that the function is estimated sufficiently correctly. It is possible to prevent presenting the estimated function when there is not, and even though the estimation is sufficiently correct, the user feels at a loss in the choice of the choice, but the choice is wasted The act of repeating can be suppressed. As a result, it is possible to provide effective support for selection of user options.

  Further, the possibility-based estimation unit updates the possibility score of each function every time input options are sequentially acquired by the option acquisition unit, and the purpose estimation device further includes an input option by the option acquisition unit. Are sequentially acquired, a time measuring unit that measures an elapsed time from the time when the option acquisition unit last acquired an input option, a presentation unit that presents a function estimated by the estimation unit, and the time When the elapsed time measured by the measuring means is longer than the fourth threshold and there is a possibility score larger than the fifth threshold among the possibility scores of the respective functions, the possibility-based estimation means It is good also as providing the presentation timing determination means which makes the said presentation means show the estimated function.

  For example, when the user is confused about selection of an option, the elapsed time from the last selection of the option becomes longer. Therefore, in the present invention, when the elapsed time is longer than the fourth threshold and any probability score is larger than the fifth threshold, that is, the user is confused about selecting an option, The function is presented when the function is already estimated sufficiently. As a result, when the user is confused about selection of options, a function that is sufficiently estimated as the user's purpose can be presented at an appropriate timing without receiving a request for assistance from the user. Moreover, it is possible to prevent a user from being confused and confused by presenting a function that is unlikely as a user's purpose.

  In addition, the purpose estimation device may further share the subgoal when there are a plurality of function names having a common subgoal among the function names of the plurality of functions estimated by the possibility-based estimation unit. It is good also as providing the presentation means to show collectively the several function name to do.

  Thereby, since each function name is categorized and presented for each subgoal focused by the user, the user can easily find the target function. That is, in the static menu (list of choices) prepared in advance by the designer of the operation target device, the user easily finds an option that leads to the target function due to a shift in the menu concept between the user and the designer. However, it can be easily found by categorizing each function for each subgoal as described above.

  The present invention can be realized not only as such an object estimation device but also as a method, a program, and a storage medium for storing the program.

  The purpose estimation device of the present invention has an effect that a function intended by a user among a plurality of functions of an operation target device can be appropriately estimated.

  Hereinafter, embodiments according to the present invention will be described with reference to the drawings.

(Embodiment 1)
FIG. 1 is a configuration diagram of a purpose estimation device according to Embodiment 1 of the present invention.

  The purpose estimation apparatus 100 according to the present embodiment can appropriately estimate a function intended by the user among a plurality of functions of the operation target device, and can sufficiently support the user's operation on the operation target device. It is. The purpose estimation apparatus 100 includes an input unit 101, an output unit 103, an input / output control unit 107, a device control unit 109, an operation history storage unit 108, an operation support determination unit 114, a purpose estimation unit 113, a subgoal estimation unit 110, A function expression storage unit 111 and a word concept storage unit 112 are provided.

  In the present embodiment, the subgoal estimation unit 110 is configured as a subgoal specifying unit, the purpose estimation unit 113 is configured as an estimation unit, and the function expression storage unit 111 is configured as a functional element storage unit.

  Such a purpose estimation device 100 is separated into, for example, an input unit 101 and a purpose estimation device body composed of components other than the input unit 101. And the objective estimation apparatus main body is equipped with the operation target apparatus. The input unit 101 is configured as, for example, a terminal (hereinafter simply referred to as a remote controller) for remotely controlling the operation target device and the purpose estimation apparatus main body.

  The input / output control unit 107 controls each component of the purpose estimation apparatus body based on a signal from the input unit 101.

  The input unit 101 receives input from the user. The input unit 101 also includes an option input unit 102 that accepts an option on the menu selected by a user operation as an input option, and a support request input unit 106 that accepts a support request from the user. The support request input unit 106 receives a support request by selecting a help button attached to the remote controller, a button with “?” Displayed on the output unit 103, or the like. Then, the input unit 101 transmits an input signal 121 indicating the received input option or the received support request to the input / output control unit 107.

  In this embodiment, the option input unit 102 is configured as an option acquisition unit that acquires an option selected by the user from a plurality of options, and the support request input unit 106 requests a support from the user. It is configured as a support request receiving means for receiving.

  The output unit 103 acquires the output signal 122 from the input / output control unit 107 and presents information indicated by the output signal 122 to the user. The output unit 103 includes an option presenting unit 104 that displays a plurality of options, and an operation support presenting unit 105 that displays support information. The support information is information for supporting the user's operation on the operation target device, and is information indicating the function intended by the user estimated by the purpose estimation device 100.

  In the present embodiment, the operation support presentation unit 105 is configured as a presentation unit that presents the function estimated by the purpose estimation unit 113.

  The operation history storage unit 108 acquires the input signal 121 indicating the input option accepted by the option input unit 102 as the selection signal 123 from the input / output control unit 107 and stores the input option indicated by the selection signal 123. When there is an operation support request, that is, when the input signal 121 indicating the support request accepted by the support request input unit 106 is output to the input / output control unit 107, the operation history storage unit 108 accumulates so far. A history signal 133 indicating an operation history that is an input option that has been selected is output to the subgoal estimation unit 110 via the input / output control unit 107.

  The device control unit 109 controls an operation target device (for example, a DVD recorder). The device control unit 109 acquires a device control signal 124 from the input / output control unit 107. The device control unit 109 controls the operation target device according to the function indicated by the device control signal 124. That is, the input / output control unit 107 specifies a function to be executed by the operation target device based on the input signal 121 acquired from the input unit 101, and outputs a device control signal 124 indicating the function to the device control unit 109. To do. In addition, when the function is executed by the operation target device, the device control unit 109 outputs a device control signal 124 indicating that the function is executed to the input / output control unit 107.

  The word concept storage unit 112 stores information that defines a word as a vector in order to calculate the similarity (score) between words. Such information can be created by a method such as latent semantic analysis (LSA), and is output to the subgoal estimation unit 110 as the word concept signal 127. That is, the word concept signal 127 indicates a vector representing the word for each of a plurality of words. Here, the LSA is created based on text information acquired from, for example, a manual of the operation target device or a Web (World Wide Web) related thereto.

  The function expression storage unit 111 stores names (function names) representing all functions of the operation target device in association with partial elements such as words included in the names. In the present embodiment, subelements (elements) are described as words. However, subelements may be phrases such as morphemes that are more detailed units or noun phrases or verb phrases that are larger units. . The information stored in the function expression storage unit 111 is output to the subgoal estimation unit 110 and the purpose estimation unit 113 as a function expression signal 128.

  The subgoal estimation unit 110 is based on the history signal 133 output from the input / output control unit 107, the word concept signal 127 obtained from the word concept storage unit 112, and the function expression signal 128 obtained from the function expression storage unit 111. For each input option indicated by the history signal 133, subgoals for all functions indicated by the function expression signal 128 are estimated. Then, the subgoal estimation unit 110 outputs a subgoal estimation result signal 125 indicating the subgoal to the input / output control unit 107.

  Here, the definition of the subgoal will be described. For example, the function name “search by person name and schedule program recording” can be divided into two parts, such as “search by person name” and “schedule recording program”, “by name”, “search by name” It is divided into four parts such as “Yes”, “Program”, and “Reserve Recording”. Thus, the function name of the operation target device is divided into a plurality of parts. Of these plural parts, the part most relevant to the input option is called a subgoal.

  In other words, a subgoal is a part of a plurality of subelements (elements) constituting a character string of a function name that is most related to the input option, that is, has the highest similarity with the input option. An element or a set of subelements. Note that the entire function name may be treated as one subgoal.

  The purpose estimation unit 113 finally estimates the user's target function based on the subgoal estimation result signal 125 output from the input / output control unit 107 and the function expression signal 128 obtained from the function expression storage unit 111. The result is output to the input / output control unit 107 as a purpose estimation result signal 130.

  The operation support determination unit 114 determines an operation support method based on the purpose estimation result signal 130 obtained through the input / output control unit 107, and outputs an operation support signal 131 for operation support.

  FIG. 2 is an explanatory diagram for explaining options and functions.

  The operation target device is, for example, a DVD recorder having a menu selection type interface.

  Generally, menus for devices such as DVD recorders and televisions are configured in a hierarchical manner, and desired functions are realized by sequentially selecting menu items (options) displayed on the screen and moving down the hierarchy. The menu is composed of a plurality of options.

  In order to manage the current state of the user, the input / output control unit 107 uses tree type (hierarchical) menu information in the operation target device, for example, as shown in FIG. This menu information indicates a node corresponding to each menu configured in a tree shape and a screen ID for identifying a display screen corresponding to each node. For example, a node written as the top menu of the first hierarchy is a node that is first selected when an option selection operation is started, and when the node is selected, a screen corresponding to the screen ID “A”; That is, the top menu 401 is displayed. When the user selects, for example, a program guide from items (options) displayed on the top menu 401, a node written as “program guide” in the second hierarchy is selected, and a screen corresponding to the screen ID “F”; That is, the menu 402 is displayed.

  That is, the option presenting unit 104 first displays the top menu 401 of the first hierarchy indicating the options “broadcast program, playback navigation, G code, etc.”. The option input unit 102 outputs to the input / output control unit 107 an input signal 121 indicating an option (input option) selected according to the operation by the user, for example, “program guide”. When the input / output control unit 107 acquires the input signal 121, the input / output control unit 107 specifies a second-level node corresponding to the input option “program guide” indicated by the input signal 121 from the menu information. Then, the input / output control unit 107 causes the option presenting unit 104 to display a menu 402 indicating the option “terrestrial analog, BS, CS” corresponding to the node in the second hierarchy. Then, the option input unit 102 outputs an input signal 121 indicating an option (input option) selected according to an operation by the user, for example, “CS”, to the input / output control unit 107. In this way, each time an option is sequentially selected, the functions to be executed by the operation target device are narrowed down from all the functions F (1) to F (N) of the operation target device. Function is specified.

  FIG. 3 is a diagram illustrating an appearance of the input unit 101.

  As described above, the input unit 101 is configured as a remote controller, and includes an option input unit 102 having a menu button 506, a cursor key 502, a back button 504, and an enter button 503, and a support request input unit 106 having a help button 507. I have. The input unit 101 has a power button 505.

  That is, the user presses the power button 505 of the input unit 101 illustrated in FIG. 3 to turn on the operation target device. Then, the user presses the menu button 506 and first displays a top menu 401 as shown in FIG. Thereafter, the user uses the cursor key 502 of the input unit 101 shown in FIG. 3 to move the cursor to a desired item (option) in the top menu 401 and presses the enter button 503. As a result, the display target node in the menu information moves to the lower layer side, and the displayed screen transitions to the next screen.

  For example, after the top menu 401 as shown in FIG. 2 is displayed, the user presses the cursor key 502 four times to move the cursor to the option “program guide”, and presses the enter button 503. As a result, the menu 402 defined in the node “program guide” in the second hierarchy in the menu information shown in FIG. 2 is displayed. In addition, when the user presses the return button 504, the node to be displayed in the menu information moves up to the upper layer side, and the top menu 401 defined in the first layer node “top menu” is displayed again.

  In addition, for example, when the user wants to view a desired program from the currently broadcast CS broadcast program, the user performs the same operation to perform “top menu” → “program guide” → “CS” → Change the displayed menu in the order of “current time” → “program list”. As a result, a menu 403 showing a list of CS broadcast programs currently being broadcast is displayed. The user can view the program by selecting the program (option) to be viewed from the menu “program list”.

  FIG. 4 is a diagram illustrating information stored in the function expression storage unit 111.

  The function expression storage unit 111 stores a function expression database 601 shown in FIG. The function expression database 601 includes a function ID for identifying a function, a name (function name) representing the function, and a function name for each function in order to estimate a function that is a user's operation purpose. A partial element and the number of elements are shown.

  Specifically, an explanatory text described in a manual or the like is used as the function name. In addition, as the partial element (element), a part corresponding to a noun or a verb extracted by performing morphological analysis on the function name (description) is used. Also, paying attention to the fact that any function expression (function name) can be described in the form of “to do”, the function name is the operation object that is the “to” part and the “to” part. Operational purposes and other elements may be used as partial elements. In this way, grammatical units may be used for the subelements.

  For example, when the function name (description) is “confirm the contents of the recorded program”, the subelements of the function name are “recording” and “program” extracted by performing morphological analysis on the function name. , “Content” and “confirmation”. In this case, “recording, program, contents” that is the operation target and “confirmation” that is the operation purpose may be used as the partial elements of the function name “confirm the contents of the recorded program”.

  Such a function expression database 601 is output as a function expression signal 128 to the subgoal estimation unit 110 and the purpose estimation unit 113.

  That is, in the present embodiment, the function expression storage unit 111 stores a plurality of elements included in the character string of the function name of each function as an element group for each function of the operation target device.

  FIG. 5 is a diagram illustrating an example of information stored in the word concept storage unit 112.

  The word concept storage unit 112 stores a word concept database 112a. The word concept database 112a shows, for each word, the word and a vector expressing the word. Each word is expressed as a vector in a semantic space constructed by a technique such as LSA. That is, the vector indicated by the word concept database 112a is a vector on a semantic space constructed by a technique such as LSA. Such a word concept database 112 a is output to the subgoal estimation unit 110 as the word concept signal 127.

  FIG. 6 is a block diagram showing the functional configuration of the subgoal estimation unit 110 in detail.

  The subgoal estimation unit 110 includes an input interpretation unit 201, a neighborhood element identification unit 202, a derivation unit 203, and an estimation unit 204. The input interpretation unit 201 expands expressions that may be interpreted as input choices (hereinafter referred to as input interpretation), and in order to prevent the neighboring element specifying unit 202 from being more or less than necessary, the interpretation of each input choice Identify the important subelements that originally make up the function. Further, a candidate as a subgoal is generated by combining the subelements identified by the derivation unit 203, and the most likely subgoal is estimated by the estimation unit 204.

  Upon receipt of the history signal 133, the input interpretation unit 201 enumerates expressions (hereinafter referred to as input interpretations) that are possible interpretations of the input options for each input option indicated by the history signal 133. An input interpretation signal 211 is output to the neighboring element specifying unit 202.

  FIG. 7 is a diagram showing the contents of the history signal 133.

  In the history signal 133, for example, the option “HDD / DVD / SD switching” is selected as an input option by the user's first selection operation, and the option “program guide” is selected as the input option by the second selection operation. This indicates that the option “return” has been selected as the input option by the third selection operation.

  FIG. 8 is a diagram illustrating input interpretation interpreted from input options.

  For example, the input interpretation unit 201 identifies (interprets) “HDD”, “DVD”, “SD”, “switch”, “HDD / DVD”, and the like as the input interpretation of the input option “HDD / DVD / SD switch”. To do. The input interpretation unit 201 then outputs an input interpretation signal 211 indicating a plurality of input interpretations for each input option as shown in FIG.

  Such an input interpretation is an element constituting the name of an input option, a set of a plurality of elements, or the like. The elements are words, phrases, morphemes and the like.

  That is, in the present embodiment, the subgoal estimation unit 110 divides a character string representing the input option acquired by the option input unit 102 into a plurality of elements, thereby generating the plurality of elements as input interpretations. An interpretation unit 201 is provided.

  Based on the input interpretation signal 211, the function expression signal 128, and the word concept signal 127, the neighborhood element specifying unit 202, for each input option and function, each of a plurality of input interpretations for the input option and a plurality of functions for the function The similarity of semantic content with each of the subelements is calculated as a score. That is, the neighborhood element specifying unit 202 calculates the inter-vector distance (for example, cosine distance) between the vector representing the input interpretation and the vector representing the partial element as a score between the input interpretation and the partial element. Further, the neighborhood element identification unit 202 identifies an important partial element (hereinafter referred to as a neighborhood element) among the partial elements of the function by comparing the calculated score with a threshold for each input option and function. Then, the neighboring element identification unit 202 outputs a neighboring element signal 212 indicating the identified neighboring element to the derivation unit 203.

  FIG. 9 is a diagram illustrating scores of input interpretation and partial elements.

  For example, the neighborhood element specifying unit 202 includes each of the input interpretation “HDD, DVD, SD,...” For the input option “HDD / DVD / SD switching” and the subelement “for confirming the contents of the recorded program”. The score between each of “recording, program, contents, confirmation” is calculated. Specifically, the neighborhood element specifying unit 202 calculates the score between the input interpretation “HDD” and the subelement “recording” as “0.26”.

  The neighborhood element specifying unit 202 also calculates a score similar to the above between the input option “HDD / DVD / SD switching” and another function “switching the audio of a recorded program”. Further, if the input interpretation signal 211 includes an input option “program guide” other than the input option “HDD / DVD / SD switching”, the neighboring element specifying unit 202 determines that the input option “program guide” and the function “ A score similar to that described above is also calculated during the period of “confirming the contents of the recorded program”.

  Further, as shown in FIG. 9, the neighborhood element specifying unit 202 uses the score and threshold value calculated for the input option “HDD / DVD / SD switching” and the function “confirm the contents of the recorded program”. By using it, an important subelement of the subelement “recording, program, content, confirmation” of the function “confirming the contents of the recorded program” is specified as a neighboring element. Specifically, when the threshold value is 0.4, the neighboring element identification unit 202 identifies the partial elements “recording” and “program” having a score of 0.4 or more as neighboring elements. Thereby, it is possible to omit subelements such as “content” and “confirmation” that are considered to be difficult to identify from the input option “HDD / DVD / SD switching”.

  Upon obtaining the neighborhood element signal 212, the derivation unit 203 derives a subgoal candidate for each input option and function based on the neighborhood element indicated by the neighborhood element signal 212. In other words, the deriving unit 203 derives a neighboring element indicated by the neighboring element signal 212 as a subgoal candidate and derives a subgoal candidate by combining a plurality of neighboring elements. For example, as described above, when the neighborhood element signal 212 indicates the neighborhood elements “recording” and “program”, the deriving unit 203 derives “recording”, “program”, and “recording program” as subgoal candidates.

  Then, the derivation unit 203 outputs the candidate signal 213 indicating the subgoal candidate derived as described above to the estimation unit 204.

  Upon obtaining the candidate signal 213, the estimating unit 204, based on the candidate signal 213 and the word concept signal 127, for each input option and function, each of a plurality of input interpretations for the input option and a plurality of the corresponding function The score between each of the subgoal candidates is calculated. Note that the estimation unit 204 calculates a score in the same procedure as that of the neighborhood element specifying unit 202. Furthermore, the estimation unit 204 estimates (specifies) a subgoal candidate corresponding to the maximum score among the calculated scores as a subgoal for each input option and function. And the estimation part 204 outputs the subgoal estimation result signal 125 which shows the subgoal estimated for every input choice and function, and its maximum score. The maximum score is hereinafter referred to as a subgoal score.

  FIG. 10 is a diagram illustrating scores of input interpretation and subgoal candidates.

  For example, the estimator 204 uses each of the input interpretations “HDD, DVD, SD,...” For the input option “HDD / DVD / SD switching” and the subgoal candidate “recording, check the content of the recorded program”. The score between each of “program and recorded program” is calculated. Specifically, the estimation unit 204 calculates the score between the input interpretation “HDD” and the subgoal candidate “recorded program” as “0.27”.

  Further, the estimation unit 204 selects the subgoal candidate “recorded program” corresponding to the maximum score “0.45” among the scores calculated in this way, the input option “HDD / DVD / SD switching”, and the function “recording”. It is estimated as a subgoal for “confirm the contents of the program that has been played”. Then, the estimation unit 204 outputs a subgoal estimation result signal 125 indicating the subgoal “recorded program” and the maximum score (subgoal score) “0.45” of the subgoal.

  Also, the estimation unit 204 estimates a subgoal by calculating a score between the input option “HDD / DVD / SD switching” and another function “switching the audio of a recorded program” in the same manner as described above. Further, if the input interpretation signal 211 includes an input option “program guide” other than the input option “HDD / DVD / SD switching”, the estimation unit 204 uses the input option “program guide” and the function “recorded”. Also during “confirm the contents of the program”, the score is calculated in the same manner as described above to estimate the subgoal. By deciding the range of functions that can be specified according to each input by the above method, the estimation was based on all the features of the input choices and functions, so it was more than necessary. The problem can be solved.

  FIG. 11 is a diagram illustrating the contents of the subgoal estimation result signal 125.

  The subgoal estimation result signal 125 indicates a subgoal for each function and a subgoal score calculated for the subgoal for each input option. For example, when acquiring the history signal 133 as shown in FIG. 7, the subgoal estimation unit 110 outputs a subgoal estimation result signal 125 as shown in FIG. In FIG. 11, double-ended arrows indicate the range of subgoals, and the numbers above the double-ended arrows indicate subgoal scores calculated for the subgoals.

  Specifically, as shown in FIG. 11, the subgoal estimation result signal 125 corresponds to the input option “HDD / DVD / SD switching”, the subgoal “recording program” of the function with the function ID “1”, and the subgoal of the subgoal. A subgoal score “0.45” is shown, a subgoal “switch recorded program audio” of the function with the function ID “2”, and a subgoal score “0.62” of the subgoal. Further, the subgoal estimation result signal 125 indicates the subgoal “recorded program” of the function having the function ID “1” and the subgoal score “0.87” of the subgoal for the input option “program guide”. The subgoal “recorded program” of the function with the function ID “2” and the subgoal score “0.87” of the subgoal are shown.

  FIG. 12 is a flowchart showing the operation of the subgoal estimation unit 110.

  For each selection operation from the first time (t = 1) to the last T time (t = T) indicated by the history signal 133, the subgoal estimation unit 110 calculates the input option selected by the selection operation and each function. Estimate subgoals and subgoal scores between.

  Specifically, first, the input interpretation unit 201 of the subgoal estimation unit 110 initializes a count number t indicating the order of selection operations by the user to t = 1, and specifies a function that is a subgoal estimation target. The count number i is initialized to i = 1 (step S901).

  Subsequently, the input interpretation unit 201 determines whether or not subgoals have been estimated for all input options indicated by the history signal 133 (step S902). For example, when the number of all input options indicated by the history signal 133 is M, the input interpretation unit 201 determines whether or not the count number t> M is satisfied, thereby determining all the input options. It is determined whether or not estimation has been performed.

  Here, when the input interpretation unit 201 determines that estimation has been performed for all input options (Y in step S902), the subgoal estimation process ends, and estimation is performed for all input options. If it is determined that it is not present (N in step S902), an input interpretation for the input option selected at the t-th time is generated (step S903). For example, the input interpretation unit 201 performs morphological analysis on the words of the input options, and generates all combinations of the obtained morphemes as input interpretations. Specifically, as shown in FIG. 8, the input interpretation unit 201 responds to the input option “HDD / DVD / SD switching” with four morphemes “HDD”, “DVD”, “ A total of 15 types of input interpretations consisting of combinations of “SD” and “switch” are generated.

  Next, the input interpretation unit 201 determines whether or not subgoal estimation has been performed between the input option selected at the t-th time and all the functions prepared in the operation target device (step S904). For example, when the number of functions prepared in the operation target device is N, the input interpretation unit 201 estimates whether all the functions are satisfied by determining whether or not the count number i> N is satisfied. It can be determined whether or not. Here, when the input interpretation unit 201 determines that the estimation has been performed with respect to all the functions (Y in step S904), the input interpretation unit 201 increases the count number t by 1 (step S905), and the process from step S902 is performed. Repeat the process.

  On the other hand, when the input interpretation unit 201 determines that estimation has not been performed with respect to all functions (N in step S904), the input interpretation unit 201 selects the input option selected for the t-th time with respect to the neighboring element identification unit 202. Then, the neighboring elements between the functions specified by the count number i are specified (step S906).

  That is, the neighboring element specifying unit 202 extracts, as neighboring elements, a partial element that has a deep relationship with the t-th input option among a plurality of partial elements that express the function of the count number i.

  Specifically, as shown in FIG. 9, the neighborhood element specifying unit 202 calculates a score between each of all input interpretations for the input options and all of the partial elements constituting the function, A partial element having the above score is extracted as a neighboring element.

  Further, when calculating the score, the neighborhood element specifying unit 202 uses a vector and a part representing the input interpretation based on the vector representation in the semantic space of each word stored in the word concept storage unit 112 shown in FIG. Calculation is performed using a vector distance (for example, cosine distance) with a vector representing an element. Here, when the input interpretation is composed of a plurality of words, the neighborhood element specifying unit 202 adds the vectors representing each of the plurality of words, performs normalization, and converts the normalized vector to the above-described one. Used as a vector representing the input interpretation.

  Then, when the neighboring element specifying unit 202 calculates the score as shown in FIG. 9, among all the partial elements constituting the function, “recording” and “program” having a score equal to or higher than the threshold (0.4) Are extracted as neighboring elements. When determining this threshold value, usability evaluation or the like is performed, and it is investigated which language expression of the function the user has a clue about the option selected when the user tried to perform a certain function. The threshold is empirically determined to match the result.

  Subsequently, the deriving unit 203 derives the neighboring element identified (extracted) by the neighboring element identifying unit 202 as a subgoal candidate, and when there are a plurality of neighboring elements, further derives a combination of the plurality of neighboring elements as a subgoal candidate. (Step S907).

  Specifically, the deriving unit 203 derives, as subgoal candidates, the neighboring elements “recording” and “program” extracted in step S906 and a “recording program” obtained by combining them.

  Subsequently, the estimation unit 204 calculates a score between the subgoal candidate derived in step S907 and all the input interpretations generated in step S903. The estimation unit 204 calculates a score using a cosine distance between a vector representing an input interpretation and a vector representing a subgoal candidate. For example, the estimation unit 204 calculates a score as shown in FIG. Then, the estimation unit 204 estimates a subgoal candidate that has obtained the highest score (subgoal score) among the calculated scores as a subgoal (step S908). In the example illustrated in FIG. 10, since the score calculated between the input interpretation “DVD” and the subgoal candidate “recorded program” is the highest score, the estimation unit 204 has the score (subgoal score). "Recorded program" is estimated as a subgoal.

  When the subgoal is estimated between the input option and the function, the input interpretation unit 201 increments the count number i (step S909) and repeatedly executes the processing from step S904.

  As described above, in the present embodiment, the subgoal estimation unit 110 is different for each function of the operation target device from one or more partial elements out of an element group (a plurality of partial elements) of the function. A plurality of subgoal candidates are set, the similarity of semantic content between the input option acquired in the option input unit 102 and each of the plurality of subgoal candidates is calculated, and the subgoal candidate corresponding to the maximum similarity is subgoal As specified.

  FIG. 13 is a configuration diagram showing the functional configuration of the purpose estimation unit 113 in detail.

  The purpose estimation unit 113 includes a possibility base estimation unit 301, a certainty factor base estimation unit 302, and a switching unit 303.

  In addition, the objective estimation part 113 in this Embodiment estimates that the function with a larger subgoal similarity (subgoal score) is made into the user's objective among several functions of operation object apparatus.

  The possibility-based estimation unit 301 receives the subgoal estimation result signal 125 and the function expression signal 128, and the higher the subgoal score calculated for any subgoal included in the function, the more the function is targeted by the user. It is estimated that there is a high possibility. And possibility base estimation part 301 outputs possibility base estimation result signal 311 which shows each function and the above-mentioned possibility (possibility score) of those functions.

  Note that the estimation performed by the possibility-based estimation unit 301 is hereinafter referred to as possibility-based estimation.

  On the other hand, the certainty factor estimation unit 302 receives the subgoal estimation result signal 125 and the function expression signal 128, and among the subelements constituting the function, the subelement whose subgoal score is calculated as a subgoal or a part of the subgoal is calculated. It is presumed that the more the number is, the more the function is intended for the user. Then, the certainty factor base estimation unit 302 outputs a certainty factor base estimation result signal 312 indicating each function and the degree of the estimation (confidence factor) of those functions.

  The estimation performed by the certainty factor base estimation unit 302 is hereinafter referred to as certainty factor based estimation.

  The switching unit 303 acquires the possibility base estimation result signal 311 and the certainty base estimation result signal 312, and switches and outputs these signals. That is, the switching unit 303 outputs the certainty factor base estimation result signal 312 as the target estimation result signal 130 if the certainty factor indicated by the certainty factor base estimation result signal 312 is high, and if the certainty factor is low, the possibility base estimation result. The signal 311 is output as the purpose estimation result signal 130.

  Here, a detailed operation of the possibility base estimation unit 301 will be described.

  The possibility base estimation unit 301 acquires the function expression signal 128 and acquires the functions and subelements of the operation target device. Further, the possibility-based estimation unit 301 acquires the subgoal estimation result signal 125 and obtains the subgoal and subgoal score of each function for each input option.

  Then, the possibility-based estimation unit 301 calculates, for each subelement of the function, a maximum subgoal score from among the subgoal scores calculated between the subgoal that includes the partial element as a part or all of the subelement and each input option. The element representative subgoal score for the partial element is specified. Further, for each function, the possibility base estimation unit 301 selects the largest element representative subgoal score from among the element representative subgoal scores specified for all the subelements included in the function, and the possibility for the function. Specify as a score.

  FIG. 14 is an explanatory diagram for explaining the operation of the possibility base estimation unit 301.

  For example, the possibility base estimation unit 301 acquires the function expression signal 128 indicating the contents of the function expression database 601 in FIG. 4 and acquires the subgoal estimation result signal 125 in FIG.

  Then, the possibility base estimation unit 301 uses the function ID “1” in the selection operation from the first (t = 1) to the last T (t = T) based on the subgoal estimation result signal 125 shown in FIG. The maximum value (maximum value in the time direction) of the subgoal score calculated for each subgoal including the subelement “recording” in the function is specified as the element representative subgoal score for the subelement “recording”. Similarly, the possibility base estimation unit 301 calculates the maximum value of the subgoal score (maximum value in the time direction) calculated for each subgoal including another partial element “program” in the function with the function ID “1”. Is specified as the element representative subgoal score for the partial element “program”.

  As a result, the possibility base estimation unit 301, as shown in FIG. 14, includes the element representative subgoal score “0.87” for the partial elements “recording” and “program” in the function with the function ID “1”, the function The element representative subgoal score “0.87” for the subelements “recording” and “program” in the function of ID “2”, and the subelements “voice” and “switching” in the function of function ID “2” The element representative subgoal score “0.62” is specified. In FIG. 14, “-” indicates that the subgoal score or the element representative subgoal score is unknown.

  FIG. 14 shows not only the element representative subgoal score when a support request is received from the user after the last T-th (t = T) selection operation, but also the selection operation for each selection operation. Also shown is an element representative subgoal score when a support request is received from the user after.

  That is, if a support request is received from the user after the first selection operation (t = 1), the possibility base estimation unit 301 selects only the first time (t = 1). In the operation, the maximum value “0.45” of the subgoal score for the subelement “recording” in the function with the function ID “1” is specified as the element representative subgoal score for the subelement “recording”. Similarly, the possibility base estimation unit 301 sets the maximum value “0.45” of the subgoal score for the other partial element “program” in the function with the function ID “1” to the element representative for the partial element “program”. Specify as subgoal score.

  Also, if a support request is received from the user after the second (t = 2) selection operation has been performed, the possibility base estimation unit 301 performs the second (t = 1) to second time. In the selection operation up to (t = 2), the maximum value “0.87” of the subgoal score for the subelement “recording” in the function with the function ID “1” is set as the element representative subgoal score for the subelement “recording”. As specified. Similarly, the possibility base estimation unit 301 sets the maximum value “0.87” of the subgoal score for the other partial element “program” in the function with the function ID “1” to the element representative for the partial element “program”. Specify as subgoal score.

  FIG. 15 is an explanatory diagram for explaining another operation of the possibility base estimation unit 301.

  When the possibility-based estimation unit 301 specifies the element representative subgoal score as described above, the possibility that the largest element representative subgoal score among the element representative subgoal scores of all the partial elements in the function is possible for the function. Identify as sex score.

  In addition, the possibility-based estimation unit 301 identifies a partial element or a set thereof having the largest element representative subgoal score among all the partial elements in the function as an estimation target range for the function.

  For example, as shown in FIG. 14, the possibility base estimation unit 301 performs the function ID “1” for the selection operation from the first time (t = 1) to the last T time (t = T). The element representative subgoal score “0.87, 0.87, −, −” for the partial element “recording, program, content, confirmation” is specified.

  In this case, as shown in FIG. 15, the possibility base estimation unit 301 specifies the largest element representative subgoal score “0.87” among them as the possibility score for the function with the function ID “1”, A set of subelements “recorded program” having the largest element representative subgoal score “0.87” is specified as an estimation target range.

  Similarly, as shown in FIG. 14, the possibility base estimation unit 301 performs the function part of the function ID “2” for the selection operation from the first time (t = 1) to the Tth time (t = T). The element representative subgoal scores “0.87, 0.87, 0.62, 0.62” for the element “recording, program, sound, switching” are specified.

  In this case, as shown in FIG. 15, the possibility base estimation unit 301 specifies the largest element representative subgoal score “0.87” among them as the possibility score for the function with the function ID “2”. A set of subelements “recorded program” having the largest element representative subgoal score “0.87” is specified as an estimation target range.

  Note that FIG. 15 also shows the selection in addition to the possibility score and the estimation target range when a support request is received from the user after the last T-th (t = T) selection operation is performed, as in FIG. In each of the operations, the possibility score and the estimation target range when the support request is received from the user after the selection operation is performed are also shown.

  That is, if a support request is received from the user after the first selection operation (t = 1), the possibility base estimation unit 301, as shown in FIG. For the function of “1”, the possibility score “0.45” and the estimation target range “recorded program” for the function are specified. Also, if a support request is received from the user after the second (t = 2) selection operation has been performed, the possibility base estimation unit 301, as shown in FIG. For the function of “1”, the possibility score “0.87” and the estimation target range “recorded program” for the function are specified.

  If the possibility base estimation unit 301 determines that a subgoal (cue) is not obtained between the predetermined function and the input option based on the subgoal estimation result signal 125, the calculation is performed as described above. A numerical value corresponding to the number of input options for which no clue as described above has been obtained may be subtracted from the obtained possibility score.

  For example, the possibility base estimation unit 301, based on the subgoal estimation result signal 125 as illustrated in FIG. 11, the input option selected in the last T-th selection operation, the function ID “1”, and the function ID “2” It is judged that the clue of the subgoal has not been obtained with the function of "". Therefore, the possibility base estimation unit 301 subtracts a predetermined numerical value from the function possibility score of the function ID “1” and the function ID “2” for which no clue is obtained as a penalty.

  As described above, the purpose estimation unit 113 according to the present embodiment calculates, for each function of the operation target device, the maximum similarity among the similarities of the plurality of subgoals specified between the function and each input option. A possibility base estimation unit 301 is provided that calculates a possibility score of the function and estimates that a function having a higher possibility score among a plurality of functions of the operation target device is intended for the user.

  Here, the detailed operation of the certainty factor base estimation unit 302 will be described.

  As shown in FIG. 14, the certainty factor base estimation unit 302 performs the process from the first time (t = 1) to the last T time (t = T) for each functional subelement, as shown in FIG. 14. In the selection operation, the maximum subgoal score among the subgoal scores calculated between each subgoal including the subelement and each input option is specified as the element representative subgoal score for the subelement.

  Then, the certainty factor base estimation unit 302 uses, for each function, the certainty factor Bf (for the function) using the element representative subgoal score and (Equation 1) specified for all the partial elements in the function. f) is calculated.

In (Equation 1), L (e _i ) is an element representative subgoal score for the subelement e _i , n (f) is the number of subelements of the function f, and i is 1 to n (f ) Is an integer up to. However, when the element representative subgoal score L (e _i ) for the partial element e _i is unknown, L (e _i ) = 0.01.

Thus, the certainty factor base estimation unit 302 calculates a high subgoal score as a subgoal or a part of the subgoal among the subelements constituting the function f by performing the calculation shown in (Expression 1). The certainty factor Bf (f) indicating a larger numerical value as the number of subelements e _i increases is calculated for the function f. Bf (f) indicates a value smaller than 0.

  FIG. 16 is a diagram illustrating the certainty factor calculated by the certainty factor base estimation unit 302.

  For example, as shown in FIG. 14, the certainty factor base estimation unit 302 performs the function ID “1” for the selection operation from the first time (t = 1) to the last T time (t = T). The element representative subgoal score “0.87, 0.87, −, −” for the partial element “recording, program, content, confirmation” is specified.

In this case, the certainty factor base estimation unit 302 has an element representative subgoal score L (e _i ) = “0.87” for each partial element e _i (i = 1 to 4) in the function f with the function ID “1”. 0.87, 0.01, 0.01 "and n (f) = 4 are substituted into (Equation 1). As a result, the certainty factor base estimation unit 302 calculates the certainty factor Bf (f) for the function f with the function ID “1” as “−1.03” as illustrated in FIG. 16.

  Similarly, as shown in FIG. 14, the certainty factor base estimation unit 302 performs the function with the function ID “2” for the selection operation from the first time (t = 1) to the last T time (t = T). Element representative subgoal scores “0.87, 0.87, 0.62, 0.62” for the subelement “recording, program, sound, switching” are specified.

In this case, the certainty factor base estimation unit 302 uses the element representative subgoal score L (e _i ) = “0.87” for each partial element e _i (i = 1 to 4) in the function f with the function ID “2”. 0.87, 0.62, 0.62 "and n (f) = 4 are substituted into (Equation 1). As a result, the certainty factor base estimation unit 302 calculates the certainty factor Bf (f) for the function f with the function ID “2” as “−0.14” as illustrated in FIG. 16.

  Thus, the certainty factor base estimation unit 302 calculates the certainty factor of the function for all the functions.

  Note that FIG. 16 is similar to FIG. 14, in addition to the certainty factor when the support request is received from the user after the last T-th (t = T) selection operation is performed, The certainty factor is also shown when a support request is received from the user after the selection operation has been performed.

  That is, if a support request is received from the user after the first selection operation (t = 1), the certainty factor base estimation unit 302, as shown in FIG. The certainty factor Bf (f) for the function “1” is calculated as “−1.18”. Also, if a support request is received from the user after the second (t = 2) selection operation is performed, the certainty factor base estimation unit 302, as shown in FIG. The certainty factor Bf (f) for the function “1” is calculated as “−1.03”.

  As described above, the purpose estimation unit 113 in the present embodiment is specified between the function and each input option among the function element group (a plurality of partial elements) for each function of the operation target device. The function of the operation target device calculates a certainty factor that decreases as the ratio of the number of subelements not included in any of the plurality of subgoals increases and increases as the similarity (subgoal score) of the plurality of subgoals increases. Among these, a certainty factor-based estimation unit 302 that estimates that a function having a higher certainty factor is intended by the user is provided.

  FIG. 17 is a flowchart showing the operation of the purpose estimation unit 113.

  The possibility-based estimation unit 301 of the purpose estimation unit 113 first specifies the possibility score and estimation target range of each function based on the subgoal estimation result signal 125 (step S1001). That is, the possibility base estimation unit 301 estimates that a function having a higher possibility score is a function intended by the user. Such an estimation method is an estimation method in which, if a high score is calculated even for some of the partial elements constituting the function, the function may be the function intended by the user.

  Subsequently, the certainty factor base estimation unit 302 of the purpose estimation unit 113 calculates the certainty factor of each function (step S1002). That is, the certainty factor base estimation unit 302 estimates that a function having a higher certainty factor is a function intended by the user. Such an estimation method is an estimation method in which a function having more high-scoring partial elements among the partial elements constituting the function is a function intended by the user.

  For example, the certainty factor base estimation unit 302 calculates the certainty factor Bf (f) of each function based on the above (Equation 1). As a result, the function whose clue (subgoal score) is obtained for more subelements has a higher certainty factor Bf (f) (a value close to 0). On the other hand, the certainty factor Bf (f) has a lower value (a value lower than 0) as the function has no clue.

  The certainty factor base estimation unit 302 calculates a certainty factor as illustrated in FIG. 16 for each function, for example. In this case, in the function of the function ID “2” and the function of the function ID “N”, as shown in FIG. There are also clues for “switching audio” and “playback”. As a result, in the function with the function ID “2” and the function with the function ID “N”, higher certainty is calculated than the function with the function ID “1” “confirm the contents of the recorded program”. In addition, the certainty factor for a function increases from the point in time when a clue is actually obtained for the function in the process of repeated selection operations.

  Subsequently, the switching unit 303 of the purpose estimation unit 113 determines whether or not the estimation result of the certainty factor base estimation unit 302 is certain (step S1003). Specifically, the switching unit 303 determines whether or not the maximum value of the certainty calculated in step S1002 is equal to or greater than a threshold (first threshold), and determines that there is certainty if it is equal to or greater than the threshold. If it is less than the threshold, it is determined that there is no certainty.

  Here, when the switching unit 303 determines that there is confidence in step S1003 (Y in step S1003), the confidence-based estimation result, that is, the confidence-based estimation result signal 312 output from the confidence-based estimation unit 302 is used. The target estimation result signal 130 is output to the input / output control unit 107 (step S1004). On the other hand, when the switching unit 303 determines that there is no certainty in step S1003 (N in step S1003), the switching unit 303 uses the possibility-based estimation result, that is, the possibility-based estimation result signal 311 output from the possibility-based estimating unit 301. The estimation result signal 130 is output to the input / output control unit 107 (step S1005).

  For example, when a support request is received from the user after the last T-th (t = T) selection operation, the certainty factors “−1.03, −0.14,..., −0. 16 "is calculated. In this case, assuming that the threshold value is “−1.05”, the switching unit 303 determines that the maximum certainty factor “−0.14” among the above certainty factors is larger than the threshold value “−0.15”. Adopt degree-based estimation results. On the other hand, if a support request is received from the user after the first (t = T) selection operation, the certainty factor “−1.18, −0.2,. 0.9 "is calculated. In this case, if the threshold value is “−1.05”, the switching unit 303 is possible because the maximum certainty factor “−0.2” of the above certainty factors is smaller than the threshold value “−0.15”. Adopt gender-based estimation results.

  By the way, the operation support determination unit 114 acquires the purpose estimation result signal 130 from the purpose estimation unit 113 via the input / output control unit 107. Then, when the purpose estimation result signal 130 indicates a certainty factor-based estimation result, the operation support determining unit 114 displays the certain functions of the operation target device in order from the function having the highest certainty factor. Determine the operation support method. On the other hand, when the purpose estimation result signal 130 indicates the possibility-based estimation result, the operation support determination unit 114 is capable of exceeding a predetermined threshold among all functions (function names) of the operation target device. A possibility-based operation support method is determined in which a plurality of functions having sex scores are collectively displayed for each function having a common estimation target range.

  Further, the operation support determination unit 114 outputs an operation support signal 131 indicating the determined operation support method to the input / output control unit 107.

  When the input / output control unit 107 acquires the operation support signal 131 from the operation support determination unit 114, the input / output control unit 107 instructs to display the function estimated by the purpose estimation unit 113 using the operation support method determined by the operation support determination unit 114. The output signal 122 of the content to be output is output to the operation support presentation unit 105 of the output unit 103.

  When obtaining the output signal 122 as described above, the operation support presenting unit 105 displays the function estimated by the purpose estimating unit 113 as support information in accordance with the content indicated by the output signal 122.

  That is, when the output signal 122 indicates the certainty factor-based estimation result, that is, the certainty factor of each function as illustrated in FIG. 16, the operation assistance presenting unit 105 uses the certainty factor-based operation assisting method of the operation target device. Display each function. That is, the operation support presentation unit 105 displays each function (function name) of the operation target device in order from the function with the highest certainty factor. On the other hand, if the output signal 122 indicates the possibility-based estimation result, that is, the possibility score of each function and the estimation target range as shown in FIG. To display each function of the operation target device. That is, the operation support presentation unit 105 selects a plurality of functions having a possibility score equal to or higher than a predetermined threshold among all functions (function names) of the operation target device for each function having a common estimation target range. Display all together.

  FIG. 18 is a diagram illustrating function names of operation target devices displayed in the possibility-based operation support method.

  As illustrated in FIG. 18, the operation support presenting unit 105 categorizes and presents a function (function name) having a possibility score equal to or higher than a predetermined value for each same estimation target range. In addition, when displaying a plurality of categories, the operation support presenting unit 105 displays the categories in descending order of the number of functions. The category is a set of functions that share the estimation target range.

  For example, the operation support presentation unit 105 has the functions “confirm the contents of the recorded program”, “switch the sound of the recorded program”, and “reproduce the recorded program” that share the estimation target range “recorded program”. And the like are collectively displayed as one category. At this time, the operation support presentation unit 105 omits expressions after the estimation target range “recorded program” to be indented among the function names, and for example, “confirm content”, “switch audio”, and “play”. Each function name is displayed as follows.

  As described above, in the present embodiment, the menu prepared in advance by the designer of the operation target device by categorizing and presenting each function (function name) around the subgoal (estimation target range) focused on by the user is used. Since the assumption about the menu structure is different between the user and the designer, the problem that the function cannot be found well can be solved. That is, the user can find the target function at an early stage rather than selecting the target function from the list of all functions. In addition, since the user selects only a function having a high possibility score rather than simply selecting a target function from all functions of the operation target device, the user can find the target function in a short time.

  FIG. 19 is a flowchart showing the operation of the purpose estimation apparatus 100 in the present embodiment.

  First, when the user presses the power button 505 of the input unit 101, the input unit 101 outputs an input signal 121 having a content for instructing to turn on the power to the input / output control unit 107. As a result, the input / output control unit 107 outputs the input signal 121 as the device control signal 124 to the device control unit 109, and causes the device control unit 109 to turn on the power of the operation target device.

  Subsequently, the input / output control unit 107 receives an input signal 121 from the input unit 101 as to whether or not the power button 505 is pressed again to instruct to turn off the power, that is, to instruct to turn off the power. It is determined whether or not it has been output (step S801). Here, when the input / output control unit 107 determines that it has been instructed to turn off the power (Y in step S801), it instructs the device control unit 109 to turn off the power of the operation target device. The device control signal 124 is output to turn off the operation target device.

  On the other hand, when the input / output control unit 107 determines that it is not instructed to turn off the power (N in step S801), it further determines whether the user has operated the input unit 101 to select an option. (Step S802). If the input / output control unit 107 determines that no option has been selected (N in step S802), it repeatedly executes the processing from step S801. On the other hand, if the input / output control unit 107 determines that an option has been selected (Y in step S802), the input option is stored in the operation history storage unit 108 as an input option (step S803). As described above, when the input options are stored in the operation history storage unit 108, as shown in FIG. 7, the input options selected so far are accumulated in the operation history storage unit 108 as an operation history.

  Next, the input / output control unit 107 executes processing corresponding to the input option acquired in step S802, that is, the latest input option stored in the operation history storage unit 108 (step S804). For example, when the user presses the menu button 506, the input / output control unit 107 acquires an input option corresponding to the menu button 506 from the input unit 101. At this time, as shown in FIG. 2, the input / output control unit 107 determines that the input option indicates the node “top menu” in the first layer of the menu information. The input / output control unit 107 executes processing corresponding to the input option “top menu”, that is, displays the top menu 401. When the user presses the cursor key 502 and the determination button 503 to select the option “program guide” shown in the top menu 401, the input / output control unit 107 uses the selected option “program guide” as an input option from the input unit 101. get. At this time, the input / output control unit 107 determines that the input option “program guide” indicates the node “program guide” in the second layer of the menu information. As a result, the input / output control unit 107 executes processing corresponding to the input option “program guide”, that is, displays the menu 402.

  When the user presses the cursor key 502 and the determination button 503 and selects the option “major league relay” shown in the menu 403, the input / output control unit 107 inputs the selected option “major league relay” as an input option. Obtained from the unit 101. At this time, the input / output control unit 107 determines that the input option “major league relay” indicates a function of “display major league relay” based on the menu information. As a result, the input / output control unit 107 causes the operation target device to execute a process corresponding to the input option “major league relay”, that is, a function of “display major league relay”. That is, the input / output control unit 107 outputs to the device control unit 109 a device control signal 124 with a content instructing to execute the function “display major league relay”.

  Next, the input / output control unit 107 determines whether or not the function is executed in step S804 (step S805). If the input / output control unit 107 determines that the function has been executed (Y in step S805), the input / output control unit 107 executes end processing such as deleting the operation history stored in the operation history storage unit 108 (step S806). The process from step S801 is repeatedly executed.

  On the other hand, when the input / output control unit 107 determines that the function is not being executed (N in step S805), the input / output control unit 107 further determines whether or not the support request is received from the user by pressing the help button 507, that is, the support request is received. It is determined whether or not an input signal 121 indicating acceptance has been output from the support request input unit 106 of the input unit 101 (step S807).

  If the input / output control unit 107 determines in step S807 that a support request has not been accepted (N in step S807), the input / output control unit 107 repeatedly executes the processing from step S801. On the other hand, when the input / output control unit 107 determines that the support request is accepted in step S807 (Y in step S807), the input / output control unit 107 outputs a history signal 133 indicating the operation history from the operation history storage unit 108, and the history signal 133 is output. Output to the subgoal estimation unit 110. As a result, the input / output control unit 107 causes the subgoal estimation unit 110 to perform subgoal and subgoal score estimation for each function for each input option (step S808).

  Further, the input / output control unit 107 causes the purpose estimation unit 113 to perform estimation of the function intended by the user using the subgoal and the subgoal score (step S809). Then, the input / output control unit 107 performs the operation support by causing the operation support presenting unit 105 to present the estimated function (step S810). At this time, for example, when the number of selection operations is small and the estimation result is uncertain, that is, when the maximum value of the certainty factor of each function is less than the threshold value, the purpose estimating apparatus 100 determines the possibility score and the estimation target range. When the support based on possibility is performed, such as displaying the function (function name) based on, and conversely, if the number of selection operations is large and the estimation result is confident, that is, the confidence of each function When the maximum value of the degree is equal to or greater than the threshold value, the support based on the certainty degree is performed such that the functions are ranked and displayed in the order of the certainty degree.

  As described above, according to the present embodiment, a partial element of a function that can be estimated from the input option is specified according to the combination of the option input by the user and the function to be compared, and configured from the partial element. By estimating the user's target function in units of sub-goals, the user's target function can be estimated appropriately without making the functions inferior to each other. can do. Furthermore, in this embodiment, when sufficient clues for estimation are not obtained, some functions intended by the user are estimated and presented based on the possibility, and sufficient clues for estimation can be obtained. In such a case, based on the certainty factor, some functions intended by the user are estimated, and they are ranked and presented. As a result, more accurate support is possible.

  Here, the effect of the purpose estimation apparatus 100 in the present embodiment will be specifically described in comparison with a conventional example.

  FIG. 20 is a diagram illustrating a result estimated by a conventional purpose estimation device for explaining the effect of the purpose estimation device 100 in the present embodiment.

  For example, in the conventional purpose estimation apparatus, the operation history as shown in FIG. 7, that is, T input options are accepted by the user's selection operation from the first time (t = 1) to the last T time (t = T). Then, when a support request is received from the user, a score as shown in FIG. 20 is calculated. That is, the conventional purpose estimation apparatus has a score “0.29” for the function with the function ID “1”, a score “0.40” for the function with the function ID “2”, and a score for the function with the function ID “N”. “0.42” is calculated, and it is estimated that the higher the score, the more the user's purpose is.

  Note that FIG. 20 shows not only the score when the support request is received from the user after the last T-th (t = T) selection operation, but also the selection operation for each selection operation. The score when a support request is received from the user is also shown.

  In such a conventional purpose estimation device, the score is calculated by comparing the feature vector calculated from the operation history up to the support request with the function explanation vector created from the entire function without considering the subgoal. doing. Therefore, conventionally, there is an unnecessarily large difference between the scores calculated for each function with the function ID “1, 2, N”. In other words, each function is more or less than necessary. Such a difference (superiority) occurs even when the support request is accepted after any selection operation.

  On the other hand, according to the possibility-based estimation method, the purpose estimation apparatus 100 according to the present embodiment uses the possibility score calculated for each function with the function ID “1, 2, N” as shown in FIG. The difference between them is small. As described above, in this embodiment, it is possible to prevent the superiority or inferiority between functions from being necessary.

  In other words, in this embodiment, it is possible to appropriately determine the clue obtained from the input options by appropriately finding the subgoal that is the target range of the function estimation from the selection operation, and requesting assistance after the first selection operation. If there is, the possibility score of both the function “confirm the contents of the recorded program” of the function ID “1” and the function “reproduce the recorded program” of the function ID “N” is calculated equally. be able to. Furthermore, in the present embodiment, when the option “program guide” is selected by the user by the second selection operation and a support request is made, the function “switching the audio of the recorded program” with the function ID “2” is provided. However, a possibility score equal to the possibility score of the two functions can be calculated.

  In the conventional purpose estimating apparatus, as shown in FIG. 20, for example, the score calculated for the function “play recorded program” of the function ID “N” is selected by the T-th selection operation. The input option “Play Navi” is not reflected greatly. In other words, the conventional purpose estimation apparatus has obtained a large clue to a part of “playback” of the function “play back recorded program” for the input option “playback navigation”, but reflects the clue. Without calculating a high score for that function.

  Furthermore, since the conventional purpose estimation device averages a plurality of input options selected by the user, when the input option “program guide” is selected by the second selection operation and a support request is received from the user, the function ID The score calculated for the function of “1” is improved (increased), and the score calculated for the function of function ID “2” and function ID “N” is deteriorated (decreased).

  On the other hand, in the object estimation device 100 according to the present embodiment, in the certainty factor-based estimation method, it is determined how high a score (subgoal score) is calculated with respect to the partial elements constituting the function. Reflecting the score (confidence), the function score (confidence) is increased as the number of subelements having a high score increases.

  Therefore, as shown in FIG. 16, the purpose estimation apparatus 100 in the present embodiment has a function “recorded” with a function ID “N” for an input option “playback navigation” selected by a T-th selection operation, for example. A large clue is obtained for a part of “playing a program”, and a high score (confidence) is calculated for the function.

  Furthermore, since the purpose estimation apparatus 100 according to the present embodiment does not average a plurality of input options selected by the user, the input option “program guide” is selected by the second selection operation and the user is selected. When the support request is received from, the scores (confidence levels) calculated for the functions with the function ID “1”, the function ID “2”, and the function ID “N” are improved (increased).

(Embodiment 2)
FIG. 21 is a configuration diagram of a purpose estimation device according to Embodiment 2 of the present invention.

  The purpose estimation apparatus in the present embodiment provides support that presents a function estimated as the user's purpose even when a support request is not received from the user.

  The purpose estimation apparatus 1900 in the present embodiment includes an input unit 101, an output unit 103, an input / output control unit 107, a device control unit 109, an operation history storage unit 108, an operation support determination unit 114, a purpose estimation unit 1901, and a subgoal estimation unit. 110, a function expression storage unit 111, a word concept storage unit 112, an operation support determination unit 1902, and a time information acquisition unit 1903. That is, the purpose estimation apparatus 1900 according to the present embodiment further includes an operation support determination unit 1902 and a time information acquisition unit 1903 as compared with the purpose estimation apparatus 100 according to the first embodiment, and the purpose estimation unit of the purpose estimation apparatus 100. Instead of 113, a purpose estimation unit 1901 is provided.

  Note that, among the components included in the purpose estimation apparatus 1900 in the present embodiment, the same reference numerals as those in the first embodiment are assigned to the same constituent elements as those in the purpose estimation apparatus 100 in the first embodiment. Detailed description is omitted. In the present embodiment, the operation support determination unit 1902 is configured as a presentation timing determination unit.

  The time information acquisition unit 1903 measures the time, and outputs a time information signal 1913 indicating the time as the time measurement result to the input / output control unit 107 and the operation support determination unit 1902.

  Each time the input option is acquired from the option input unit 102 of the input unit 101, the input / output control unit 107 specifies the time when the input option is acquired, that is, the time when the user selects the input option (selection time). Further, every time an input option is acquired, the input / output control unit 107 stores the input option in the operation history storage unit 108 and also includes an operation history including the input option stored in the operation history storage unit 108. As a history signal 133 and output to the subgoal estimation unit 110.

  Further, when the input / output control unit 107 acquires the possibility-based estimation result signal 311 and the certainty-based estimation result signal 312 from the purpose estimation unit 1901 and specifies the selection time based on the time information signal 1913, The selected times are collectively output to the operation support determination unit 1902 as a determination input signal 1911.

  FIG. 22 is a diagram illustrating the contents of the determination input signal 1911.

  For example, as illustrated in FIG. 22, the determination input signal 1911 indicates a selection time, a possibility score of each function at the selection time, and a certainty factor for each input option selection operation. Specifically, the determination input signal 1911 includes a selection time “10:00” in the first selection operation (t = 1), a function possibility score “0.45” and a certainty factor “−” of the function ID “1”. 1.18 ”, function possibility score“ 0.62 ”of function ID“ 2 ”, and certainty factor“ −0.2 ”.

  In addition, when the input / output control unit 107 acquires the input signal 121 indicating that the support request has been received from the support request input unit 106 of the input unit 101, the input / output control unit 107 uses the input signal 121 as the support request signal 121a. Output to.

  Whenever an input option is selected by the user, the subgoal estimation unit 110 acquires an operation history including the input option as the history signal 133, and estimates a subgoal and subgoal score between the input option and each function. The estimation result is output as a subgoal estimation result signal 125 to the purpose estimation unit 1901 via the input / output control unit 107.

  Every time the goal estimation unit 1901 acquires the subgoal estimation result signal 125, that is, every time an input option is selected by the user, the goal estimation unit 1901 performs possibility-based and confidence-based estimation based on the subgoal estimation result signal 125. The possibility-based estimation result signal 311 and the certainty factor-based estimation result signal 312 indicating those estimation results are output to the input / output control unit 107 without switching.

  That is, as shown in FIGS. 14 to 16, the purpose estimation unit 1901 in the present embodiment performs element representative subgoal scores for each function based on the input options selected so far, every time a selection operation is performed. Is calculated, the possibility score and the estimation target range are specified, and the certainty factor is calculated.

  When the operation support determination unit 1902 acquires the determination input signal 1911 from the input / output control unit 107, the operation support determination unit 1902 determines whether or not it is time to perform operation support for the user based on the determination input signal 1911. Then, when the operation support determination unit 1902 determines that it is time to perform the operation support, one of the possibility base estimation result signal 311 and the certainty base estimation result signal 312 is input as the operation support determination result signal 1912. Output to the output control unit 107. The input / output control unit 107 that has acquired the operation support determination result signal 1912 outputs the operation support determination result signal 1912 to the operation support determination unit 114 as the purpose estimation result signal 130.

  Similar to the first embodiment, such purpose estimation apparatus 1900 presents the function estimated as the user's purpose at the timing when the support request is received from the user, and determines that the user is in a confused state. The operation support is executed so as to present the estimated function at the timing when the estimated function and the function intended by the user are appropriately estimated.

  That is, the purpose estimation apparatus 1900 in the present embodiment performs operation support at the following three timings.

First support timing: Timing at which the purpose estimation device 1900 determines that the user cannot perform a new selection operation and is worried for a predetermined period of time (is in a confused state). Second support timing: The purpose estimation device 1900 Timing at which the user's target function is estimated with high confidence while the user is performing an option selection operation Third support timing: Timing at which the purpose estimation device 1900 receives a support request from the user

  It should be noted that the objective estimation device 1900 provides support such as presenting a possibility-based estimation result when providing support at the first support timing, and provides the confidence-based estimation result when providing support at the second support timing. Provide support as presented. In addition, when providing support at the third support timing, the purpose estimation apparatus 1900 provides support that presents either one of the possibility-based estimation result or the certainty-based estimation result, as in the first embodiment. Do.

  FIG. 23 is a configuration diagram showing in detail the functional configuration of the purpose estimation unit 1901 in the present embodiment.

  Similar to the purpose estimation unit 113 of the first embodiment, the purpose estimation unit 1901 includes a possibility base estimation unit 301 and a certainty factor base estimation unit 302.

  The possibility base estimation unit 301 acquires the subgoal estimation result signal 125, performs possibility based estimation based on the subgoal estimation result signal 125, and outputs the estimation result as a possibility base estimation result signal 311.

  The certainty factor base estimation unit 302 acquires the subgoal estimation result signal 125, performs certainty factor based estimation based on the subgoal estimation result signal 125, and outputs the estimation result as the certainty factor based estimation result signal 312.

  FIG. 24 is a configuration diagram showing in detail the functional configuration of the operation support determination unit 1902 in the present embodiment.

  The operation support determination unit 1902 includes a certainty factor change calculation unit 2801, an elapsed time identification unit 2802, and a determination output unit 2803.

  The certainty factor change calculation unit 2801 acquires the determination input signal 1911 and calculates the change amount of the certainty factor of each function indicated by the determination input signal 1911. That is, the certainty factor change calculation unit 2801, as shown in (Expression 2), the certainty factor of the function f calculated by the last selection operation (t = T) and the previous selection operation (t = T−1). ) Is calculated as ΔBf (f), and the average value of ΔBf (f) of each function is calculated as the above-described change amount ΔBfa. Then, the certainty factor change calculation unit 2801 notifies the determination output unit 2803 of the calculated certainty factor change amount ΔBfa.

  That is, in this embodiment, the certainty factor change calculation unit 2801 calculates the difference between the average value of the certainty factor of each function for the last selection operation and the average value of the certainty factor of each function for the previous selection operation. It is calculated as ΔBfa. The certainty factor change calculation unit 2801 may calculate the difference of the moving average as the change amount ΔBfa. That is, the certainty factor change calculation unit 2801 calculates the average value of the certainty factor of each function with respect to n selection operations up to the last selection operation and the certainty factor of each function with respect to n selection operations until the previous selection operation. The difference from the average value may be calculated as the change amount ΔBfa.

  The elapsed time specifying unit 2802 acquires the determination input signal 1911 and the time information signal 1913, and specifies the elapsed time since the last selection operation by the user based on these signals. Then, the elapsed time specifying unit 2802 notifies the determination output unit 2803 of the specified elapsed time.

  The determination output unit 2803 acquires the determination input signal 1911. When the determination output unit 2803 acquires the support request signal 121a from the input / output control unit 107, the determination output unit 2803 determines that the current time is the timing at which the user should support the operation. At this time, the determination output unit 2803 performs the same processing as the switching unit 303 of the first embodiment, and either one of the possibility base estimation result signal 311 and the certainty factor base estimation result signal 312 is used as the operation support determination result signal 1912. Output as.

  When the determination output unit 2803 has not acquired the support request signal 121a, the degree of certainty change ΔBfa notified from the certainty factor change calculating unit 2801 and the elapsed time Δt notified from the elapsed time specifying unit 2802 Based on the certainty factor and the possibility score indicated by the determination input signal 1911, it is determined whether or not the current time is the timing at which the user should support the operation. At this time, the determination output unit 2803 outputs one of the possibility base estimation result signal 311 and the certainty base estimation result signal 312 as the operation support determination result signal 1912 based on the determination result.

  That is, the determination output unit 2803 determines whether the current time is the first support timing or the second support timing when the support request signal 121a is not acquired. When the determination output unit 2803 determines that the current time is the first support timing, the determination output unit 2803 outputs the possibility base estimation result signal 311 so that the estimation result indicated by the possibility base estimation result signal 311 is presented. On the other hand, when the determination output unit 2803 determines that the current time is the second support timing, the determination output unit 2803 outputs the reliability base estimation result signal 312 so that the estimation result indicated by the reliability base estimation result signal 312 is presented.

  Here, the operation of the determination output unit 2803 at the first support timing will be described in detail with reference to FIG. 25 and FIG.

  FIG. 25 is a diagram illustrating an example of a specific selection operation.

  For example, in the first selection operation (t = 1), the user selects the input option “HDD / DVD / SD switching” at the selection time “10 o'clock”, and the selection time “10: 3 sec. ”Is selected as the second selection (t = 2) selection option“ program guide ”. Thereafter, the user is worried without selecting a new input option for a while.

  FIG. 26 is an explanatory diagram for explaining the operation of the determination output unit 2803 at the first support timing.

  The determination output unit 2803 always determines whether or not the current time is the first support timing. Hereinafter, an example of a specific operation of the determination output unit 2803 when 2 seconds have elapsed since the first selection operation illustrated in FIG. 25 and when 11 seconds have elapsed since the second selection operation has been performed. Will be explained.

  First, as shown in FIG. 25, when the input option “HDD / DVD / SD switching” is selected by the first selection operation, the determination output unit 2803 of the operation support determination unit 1902 is based on the determination input signal 1911. The possibility score and the certainty factor of each function calculated by the first selection operation are acquired. Specifically, as illustrated in FIG. 26, the determination output unit 2803 has a function possibility score “0.45” and a certainty factor “−1.18” of the function ID “1”, and a function ID “2”. The function possibility score “0.62” and the certainty factor “−0.2”, and the function possibility score “0.45” and the certainty factor “−0.9” of the function ID “N” are acquired. .

  At this time, the certainty factor change calculation unit 2801 of the operation support determination unit 1902 calculates the average value “−0.76” of the certainty factor of each function by the first selection operation. In addition, since there is no previous selection operation, the certainty factor change calculation unit 2801 notifies the determination output unit 2803 of the above-described certainty factor change amount ΔBfa as indefinite (“−” in FIG. 26).

  Further, the elapsed time specifying unit 2802 of the operation support determination unit 1902 measures the elapsed time Δt after the first selection operation is performed and notifies the determination output unit 2803 of it. That is, the elapsed time specifying unit 2802 measures the elapsed time Δt by subtracting the selection time “10 o'clock” of the first selection operation from the time indicated by the time information signal 1913.

  Accordingly, until the next second selection operation is performed, the determination output unit 2803, the probability score of each function calculated by the first selection operation, the certainty degree change ΔBfa, and the elapsed time Δt. Are used to determine whether the current time is the first support timing.

  That is, the determination output unit 2803 has the elapsed time Δt longer than the time threshold value (fourth threshold value) th, the absolute value of the confidence change amount ΔBfa is smaller than the change amount threshold value θa, and the score threshold value (the fifth threshold value). In the case where there is a possibility score larger than γa, it is determined that the current time is the first support timing.

  Such discrimination is based on the following grounds.

  The determination output unit 2803 estimates that the user may be confused when the elapsed time Δt exceeds the time threshold th. However, since the user is not confused and is concerned about other things, the elapsed time Δt may increase. Therefore, the determination output unit 2803 uses the change amount ΔBfa of the certainty factor in order to avoid erroneously determining that the user is distracted by being confused. It is checked whether there is a deadlock in the selection operation.

  In other words, when the user is confused by the selection operation, the same menu is searched, and a new clue for estimating the purpose cannot be obtained. As a result, there is a tendency that the value of the certainty factor Bf (f) does not change. Therefore, if the absolute value of the certainty degree change amount ΔBfa is smaller than the change amount threshold value θa, the determination output unit 2803 is not simply concerned about something else, and the selection operation is stuck. Judge.

  Therefore, the determination output unit 2803 is confused because the user is stuck in the selection operation when the elapsed time Δt is longer than the time threshold th and the absolute value of the confidence change ΔBfa is smaller than the change threshold θa. It is determined that the user is in such a confused state as a criterion for determining the first support timing.

  Furthermore, even if a function based on the possibility score is presented at a stage where only a low possibility score is obtained for any function, the function may not be a function intended by the user.

  Therefore, the determination output unit 2803 determines that a function having a high possibility as a user's purpose is estimated when the possibility score of any function is larger than the score threshold γa, and such a function is estimated. This is used as a criterion for determining the first support timing.

  For example, the time threshold th is th = 10 seconds, the variation threshold θa is θa = 0.5, and the score threshold γa is γa = 0.6.

  In such a case, if the elapsed time Δt at the current time is Δt = 2 seconds, the determination output unit 2803 may determine that one possibility score “0.62” is larger than the score threshold γa = 0.6. Since the elapsed time Δt = 2 seconds is smaller than the time threshold th = 10 seconds and the certainty factor change amount ΔBfa is indefinite, the current time (when 2 seconds have passed since the first selection operation is performed) is the first. It is determined that it is not the support timing.

  Next, as illustrated in FIG. 25, when the input option “program guide” is selected by the second selection operation, the determination output unit 2803 of the operation support determination unit 1902 performs the second time based on the determination input signal 1911. The possibility score and the certainty factor of each function calculated by the selection operation are acquired. Specifically, the determination output unit 2803 has a function possibility score “0.87” and a certainty factor “−1.03” of the function ID “1”, and a function possibility score “0” of the function ID “2”. .87 ”and the certainty factor“ −0.14 ”, the function possibility score“ 0.87 ”and the certainty factor“ −0.53 ”of the function ID“ N ”.

  At this time, the certainty factor change calculation unit 2801 of the operation support determination unit 1902 calculates the average value “−0.56” of the certainty factor of each function by the second selection operation. Further, the certainty factor change calculation unit 2801 has an average value “−0.76” of the certainty factor of each function by the first selection operation and an average value “−0. The determination output unit 2803 is notified of the difference “−0.20” from “56” as the above-described certainty degree change ΔBfa.

  Further, the elapsed time specifying unit 2802 of the operation support determination unit 1902 measures the elapsed time Δt after the second selection operation is performed and notifies the determination output unit 2803 of it. That is, the elapsed time specifying unit 2802 measures the elapsed time Δt by subtracting the selection time “10: 3” of the second selection operation from the time indicated by the time information signal 1913.

  Therefore, the determination output unit 2803 displays the possibility score, the change amount ΔBfa of the certainty factor, and the elapsed time Δt calculated by the second selection operation until the next third selection operation is performed. By comparing with the score threshold value γa, the change amount threshold value θa, and the time threshold value th, it is determined again whether or not the current time is the first support timing.

  For example, as described above, the time threshold th is th = 10 seconds, the variation threshold θa is θa = 0.5, and the score threshold γa is γa = 0.6.

  In such a case, if the current elapsed time Δt is Δt = 11 seconds, the determination output unit 2803 indicates that the elapsed time Δt = 11 seconds is longer than the time threshold th = 10 seconds, and the certainty factor variation ΔBfa Since the absolute value “0.2” is larger than the variation threshold θa = 0.5 and the three possibility scores “0.87” are larger than the score threshold γa = 0.6, It is determined that the first support timing is 11 seconds after the selection operation is performed.

  When the determination output unit 2803 determines that the current time is the first support timing in this way, the determination output unit 2803 outputs the possibility base estimation result signal 311 as the operation support determination result signal 1912.

  As described above, the purpose estimation apparatus 1900 according to the present embodiment can be used as a user's purpose when the user has not found a target function and is worried that a new selection operation cannot be performed. Efficient support is realized by providing support that presents a function with a high probability, that is, a function with a high possibility score.

  Next, the operation of the determination output unit 2803 at the second support timing will be described in detail with reference to FIGS.

  FIG. 27 is a diagram illustrating another example of a specific selection operation.

  For example, in the first selection operation (t = 1), the user selects the input option “HDD / DVD / SD switching” at the selection time “10 o'clock”, and the selection time “10: 3 sec. ”, The input option“ program guide ”is selected as the second (t = 2) selection operation, and the input option“ reproduction navigation ”is selected as the third (t = 3) selection operation four seconds later. .

  FIG. 28 is an explanatory diagram for explaining the operation of the determination output unit 2803 at the second support timing.

  Each time the selection operation is performed by the user and the input option is selected, the determination output unit 2803 determines whether or not the current time is the second support timing with the time when the selection operation is performed as the current time.

  Hereinafter, an example of a specific operation of the determination output unit 2803 when the second and third selection operations illustrated in FIG. 27 are performed will be described.

  As illustrated in FIG. 27, when the input option “program guide” is selected by the second selection operation, the determination output unit 2803 of the operation support determination unit 1902, based on the determination input signal 1911, 2 The possibility score and the certainty factor of each function calculated by the second selection operation are acquired. Specifically, the determination output unit 2803 has a function possibility score “0.87” and a certainty factor “−1.03” of the function ID “1”, and a function possibility score “0” of the function ID “2”. .87 ”and the certainty factor“ −0.14 ”, the function possibility score“ 0.87 ”and the certainty factor“ −0.53 ”of the function ID“ N ”.

  At this time, the certainty factor change calculation unit 2801 of the operation support determination unit 1902 calculates the average value “−0.56” of the certainty factor of each function by the second selection operation, as described above. Further, the certainty factor change calculation unit 2801 has an average value “−0.76” of the certainty factor of each function by the first selection operation and an average value “−0. The determination output unit 2803 is notified of the difference “−0.20” from “56” as the above-described certainty degree change ΔBfa.

  Therefore, the determination output unit 2803 uses the certainty factor of each function calculated by the second selection operation and the change amount ΔBfa of the certainty factor as the second support timing. It is determined whether or not.

  That is, the determination output unit 2803 has a certainty factor Bf () in which the absolute value of the certainty degree change amount ΔBfa is smaller than the change amount threshold value (second threshold value) θb (θb <θa) and larger than the certainty factor threshold value γb. If there is f), it is determined that the current time is the second support timing.

  Such discrimination is based on the following grounds.

  The user strongly desires operation support as he feels that the selection operation is stuck. Further, as described above, when the user gets stuck in the selection operation, the same menu is searched, and new clues for estimating the purpose cannot be obtained. As a result, there is a tendency that the value of the certainty factor Bf (f) does not change. Therefore, the determination output unit 2803 determines that the user's selection operation is stalled when the absolute value of the certainty degree variation ΔBfa is smaller than the variation threshold θb.

  Therefore, the determination output unit 2803 determines that the user is stuck in the selection operation when the absolute value of the change amount ΔBfa of the certainty factor is smaller than the change amount threshold value θa, and determines that the user is so stuck. The second support timing is determined as a criterion.

  Furthermore, even if the function f based on the certainty factor Bf (f) is presented at a stage where only a low certainty factor Bf (f) is obtained for any function, the function f is the user's purpose. It may not be a function.

  Therefore, when the certainty factor of any function is greater than the certainty factor threshold γb, the determination output unit 2803 determines that a function having a high certainty factor is estimated as the user's purpose, and such a function is estimated. This is used as a criterion for determining the second support timing.

  For example, the variation threshold θb is θb = 0.15, and the certainty threshold γb is γb = −0.2.

  In such a case, the determination output unit 2803 has an absolute value of the certainty degree change ΔBfa = −0.20 even if one certainty degree “−0.14” is larger than the certainty degree threshold γb = −0.2. Since the value “0.2” is larger than the variation threshold θb = 0.15, it is determined that the current time (when the second selection operation is performed) is not the second support timing.

  Next, as shown in FIG. 27, when the input option “reproduction navigation” is selected by the third selection operation before the time threshold th (= 10 seconds) has elapsed since the second selection operation, the operation support determination is made. Based on the determination input signal 1911, the determination output unit 2803 of the unit 1902 obtains the possibility score and the certainty factor of each function calculated by the third selection operation. Specifically, the determination output unit 2803 has a function possibility score “0.87” and a certainty factor “−1.03” of the function ID “1”, and a function possibility score “0” of the function ID “2”. .87 ”and the certainty factor“ −0.14 ”, the function possibility score“ 0.87 ”and the certainty factor“ −0.16 ”of the function ID“ N ”.

  At this time, the certainty factor change calculation unit 2801 of the operation support determination unit 1902 calculates the average value “−0.43” of the certainty factor of each function by the third selection operation. Further, the certainty factor change calculation unit 2801 has an average value “−0.56” of the certainty factor of each function by the second selection operation and an average value “−0. The difference “−0.13” from “43” is notified to the determination output unit 2803 as the above-described certainty degree change ΔBfa.

  Therefore, the determination output unit 2803 compares the absolute value of the certainty factor of each function and the certainty factor variation ΔBfa calculated by the third selection operation with the certainty factor threshold value γb and the variation threshold value θb again. It is determined whether or not the current time when the third selection operation is performed is the second support timing.

  For example, as described above, the variation threshold θb is θb = 0.15, and the certainty threshold γb is γb = −0.2.

  In such a case, the determination output unit 2803 determines that the absolute value “0.13” of the certainty degree change ΔBfa = −0.13 is smaller than the change amount threshold θb = 0.15, and the two certainty degrees “ Since “−0.14” and “−0.16” are larger than the certainty threshold γb = −0.2, it is determined that the current time (when the third selection operation is performed) is the second support timing. .

  When the determination output unit 2803 determines that the current time is the second support timing in this way, the determination output unit 2803 outputs the certainty factor base estimation result signal 312 as the operation support determination result signal 1912.

  As described above, the purpose estimation apparatus 1900 according to the present embodiment presents a function having a high certainty degree if a high degree of certainty is obtained at a stage where the user finds a target function and gets stuck. By providing support, it is possible to reduce the search for useless functions by the user, that is, unnecessary selection operations.

  FIG. 29 is a flowchart showing the operation of the operation support determination unit 1902.

  First, the operation support determination unit 1902 acquires the determination input signal 1911 (step S100), and determines whether or not a support request has been received from the user, that is, whether or not the support request signal 121a has been acquired (step S102). .

  If the operation support determination unit 1902 determines that a support request has been received (Y in step S102), the operation support determination unit 1902 determines that the function intended by the user should be presented at the third support timing. At this time, the operation support determination unit 1902 determines whether any of the certainty factors Bf (f) calculated for each function is larger than the certainty factor threshold γ (step S104). When the operation support determination unit 1902 determines that any one of the certainty factors Bf (f) is greater than the certainty factor threshold γ (Y in step S104), the operation support determination result signal 312 is output as the operation support determination result signal 1912. If it is determined that the certainty threshold value γ or less (N in Step S104), the possibility-based estimation result signal 311 is output as the operation support determination result signal 1912 (Step S108).

  If the operation support determination unit 1902 determines that a support request has not been received in step S102 (N in step S102), the elapsed time Δt from when the determination input signal 1911 is acquired is the time threshold th. It is determined whether or not the time is longer (step S110). That is, the operation support determination unit 1902 determines whether or not the user may be confused.

  If the operation support determination unit 1902 determines that the elapsed time Δt is longer than the time threshold th (Y in step S110), is the absolute value of the certainty change ΔBfa smaller than the change threshold θa? It is determined whether or not (step S112). That is, the operation support determination unit 1902 determines whether or not there is a deadlock in the user's selection operation.

  When the operation support determination unit 1902 determines that the absolute value of the change amount ΔBfa of the certainty factor is smaller than the change amount threshold θa (Y in step S112), the operation support determination unit 1902 determines that the user is stuck in the selection operation, It is determined whether there is a possibility score larger than the score threshold γa (step S114).

  Here, if the operation support determination unit 1902 determines that there is a possibility score greater than the score threshold γa (Y in step S114), the operation support determination unit 1902 determines that the function intended by the user should be presented at the first support timing. At this time, the operation support determination unit 1902 outputs the possibility base estimation result signal 311 as the operation support determination result signal 1912 (step S116).

  Further, when the operation support determination unit 1902 determines that the elapsed time Δt is equal to or less than the time threshold th in step S110 (N in step S110), the operation support determination unit 1902 further determines whether or not the next determination input signal 1911 has been acquired (step S118). ). When the operation support determination unit 1902 determines that the next determination input signal 1911 has been acquired (Y in step S118), the reliability is determined by comparing the previous determination input signal 1911 with the acquired determination input signal 1911. Change amount ΔBfa is calculated (step S120).

  Further, the operation support determination unit 1902 determines whether or not the absolute value of the certainty degree change amount ΔBfa is smaller than the change amount threshold value θb (step S122). That is, the operation support determination unit 1902 determines whether or not there is a deadlock in the user's selection operation. Here, when the operation support determination unit 1902 determines that the absolute value of the change amount ΔBfa of the certainty factor is smaller than the change amount threshold θb (Y in step S122), the operation support determination unit 1902 determines that the user is stuck in the selection operation, and Then, it is determined whether or not there is a certainty factor Bf (f) greater than the certainty factor threshold value γb (step S124).

  Here, when the operation support determination unit 1902 determines that there is a certainty factor Bf (f) greater than the certainty factor threshold γb (Y in step S124), the function intended by the user has already been appropriately estimated, and the second It is determined that the function should be presented at the support timing. At this time, the operation support determination unit 1902 outputs the certainty factor base estimation result signal 312 as the operation support determination result signal 1912 (step S126).

  When the operation support determination unit 1902 determines in step S112 that the absolute value of the change amount ΔBfa of the certainty factor is greater than or equal to the change amount threshold value θa (N in step S112), all possibility scores are scored in step S114. When it is determined that it is equal to or less than the threshold value γa (N in step S114), or when it is determined in step S118 that the next determination input signal 1911 is not acquired (N in step S118), the change amount ΔBfa of the certainty factor in step S122. Is determined to be greater than or equal to the variation threshold θb (N in step S122), and when it is determined in step S124 that all the certainty factors Bf (f) are less than or equal to the certainty factor threshold γb (step S124). N), it is determined whether or not all the processes should be terminated (step S128). If it is determined that the operation support determination unit 1902 should not end (N in step S128), the processing from step S102 is repeatedly executed. If it is determined that the operation support determination unit 1902 should be ended (Y in step S128), all the processing ends. .

  FIG. 30 is a flowchart showing the operation of the purpose estimation apparatus 1900 in the present embodiment.

  First, when the user presses the power button 505 of the input unit 101, the input unit 101 outputs an input signal 121 having a content for instructing to turn on the power to the input / output control unit 107. As a result, the input / output control unit 107 outputs the input signal 121 as the device control signal 124 to the device control unit 109, and causes the device control unit 109 to turn on the power of the operation target device.

  Subsequently, the input / output control unit 107 receives an input signal 121 from the input unit 101 as to whether or not the power button 505 is pressed again to instruct to turn off the power, that is, to instruct to turn off the power. It is determined whether it has been output (step S200). Here, when the input / output control unit 107 determines that it has been instructed to turn off the power (Y in step S200), it instructs the device control unit 109 to turn off the power of the operation target device. The device control signal 124 is output to turn off the operation target device.

  On the other hand, when the input / output control unit 107 determines that it is not instructed to turn off the power (N in step S200), it further determines whether the user has operated the input unit 101 to select an option. (Step S202). If the input / output control unit 107 determines that no option has been selected (N in step S202), the operation support determination unit 1902 determines whether the current time is the first support timing (step S204). . At this time, the operation support determination unit 1902 determines whether or not the current time is the first support timing by performing the processing of steps S110 to S114 shown in FIG.

  If it is determined that the current time is not the first support timing (N in step S204), the input / output control unit 107 repeatedly executes the processing from step S200. On the other hand, when it is determined that the current time is the first support timing (Y in step S204), the input / output control unit 107 determines the operation support determination result signal 1912 (possibility base estimation result) output from the operation support determination unit 1902. The signal 311) is acquired, and the operation support determination result signal 1912 is output to the operation support determination unit 114 as the purpose estimation result signal 130. As a result, the operation support presentation unit 105 performs operation support by displaying the estimation result of the purpose estimation unit 1901 (step S206). In this case, the operation support presentation unit 105 displays the possibility-based estimation result by the possibility-based operation support method shown in FIG.

  If the input / output control unit 107 determines that an option is selected in step S202 (Y in step S202), the input / output control unit 107 stores the selected option in the operation history storage unit 108 as an input option (step S208). As described above, when the input options are stored in the operation history storage unit 108, as shown in FIG. 7, the input options selected so far are accumulated in the operation history storage unit 108 as an operation history.

  Next, the input / output control unit 107 executes processing according to the input option acquired in step S202, that is, the latest input option stored in the operation history storage unit 108 (step S210).

  Next, the input / output control unit 107 determines whether or not the function is executed in step S210 (step S212). If the input / output control unit 107 determines that the function has been executed (Y in step S212), the input / output control unit 107 executes end processing such as deleting the operation history stored in the operation history storage unit 108 (step S214). The process from step S200 is repeatedly executed.

  On the other hand, when the input / output control unit 107 determines that the function is not executed (N in step S212), the operation history including the input options acquired in step S202 is output as the history signal 133 from the operation history storage unit 108. The history signal 133 is output to the subgoal estimation unit 110. Thereby, the input / output control unit 107 causes the subgoal estimation unit 110 to perform estimation of the subgoal and subgoal score of each function for each input option (step S216).

  Further, the input / output control unit 107 causes the purpose estimation unit 1901 to perform estimation of the function intended by the user using the subgoal and the subgoal score (step S218).

  Here, the input / output control unit 107 causes the operation support determination unit 1902 to determine whether or not the current time is the third support timing (step S220). At this time, the operation support determination unit 1902 determines whether or not the current time is the third support timing by executing the process of step S102 of FIG.

  When the operation support determination unit 1902 determines that the current time is the third timing (Y in step S220), the input / output control unit 107 acquires the operation support determination result signal 1912 from the operation support determination unit 1902, and The operation support determination result signal 1912 is output to the operation support determination unit 114 as the purpose estimation result signal 130. As a result, the operation support presentation unit 105 performs operation support by displaying the estimation result of the purpose estimation unit 1901 (step S206).

  On the other hand, when the operation support determination unit 1902 determines that the current time is not the third timing (N in step S220), the operation support determination unit 1902 further determines whether the current time is the second support time (step S222). At this time, the operation support determination unit 1902 determines whether or not the current time is the second support timing by executing the processes of steps S122 and S124 of FIG.

  When the operation support determination unit 1902 determines that the current time is the second support timing (Y in step S222), the input / output control unit 107 receives the operation support determination result signal 1912 (confidence-based basis) from the operation support determination unit 1902. The estimation result signal 312) is acquired, and the operation support determination result signal 1912 is output to the operation support determination unit 114 as the purpose estimation result signal 130. As a result, the operation support presentation unit 105 performs operation support by displaying the estimation result of the purpose estimation unit 1901 (step S206). In this case, the operation support presentation unit 105 displays a certainty factor-based estimation result.

  On the other hand, when the operation support determination unit 1902 determines that the current time is not the second support timing (N in step S222), the input / output control unit 107 repeatedly executes the processing from step S200.

  Thus, in this embodiment, even when the user does not press the help button 507 provided in the support request input unit 115 of the input unit 101, when the user is confused about the selection operation of the option, it is correctly estimated. Functions can be presented to the user.

  The object estimation apparatus of the present invention can be used for an interface for specifying a function by selecting an option presented on the screen. For example, a television having a hierarchical interface, a DVD player, a DVD recorder, It can be applied to mobile phones. Further, it is also useful as an operation support device for a device such as a personal computer having an interface in which a function in the menu cannot be seen unless a menu is selected, such as a pull-down menu.

It is a block diagram of the objective estimation apparatus in Embodiment 1 of this invention. It is explanatory drawing for demonstrating the option and function same as the above. It is a figure which shows the external appearance of an input part same as the above. It is a figure which shows the information stored in the function expression storage part same as the above. It is a figure which shows an example of the information stored in the word concept memory | storage part same as the above. It is a block diagram which shows the functional structure of a subgoal estimation part same as the above in detail. It is a figure which shows the content of the history signal same as the above. It is a figure which shows the input interpretation interpreted from the input choice same as the above. It is a figure which shows the score of an input interpretation same as the above, and a partial element. It is a figure which shows the score of an input interpretation same as the above and a subgoal candidate. It is a figure which shows the content of the subgoal estimation result signal same as the above. It is a flowchart which shows operation | movement of a subgoal estimation part same as the above. It is a block diagram which shows the functional structure of a purpose estimation part same as the above in detail. It is explanatory drawing for demonstrating operation | movement of a possibility base estimation part same as the above. It is explanatory drawing for demonstrating other operation | movement of a possibility base estimation part same as the above. It is a figure which shows the reliability calculated by the reliability base estimation part same as the above. It is a flowchart which shows operation | movement of the objective estimation part same as the above. It is a figure which shows the function name of the operation target apparatus displayed by the possibility-based operation support method same as the above. It is a flowchart which shows operation | movement of the objective estimation apparatus same as the above. It is a figure which shows the result estimated by the conventional objective estimation apparatus for demonstrating the effect of the objective estimation apparatus same as the above. It is a block diagram of the objective estimation apparatus in Embodiment 2 of this invention. It is a figure which shows the content of the determination input signal same as the above. It is a block diagram which shows the functional structure of a purpose estimation part same as the above in detail. It is a block diagram which shows the functional structure of an operation assistance determination part same as the above in detail. It is a figure which shows an example of the specific selection operation same as the above. It is explanatory drawing for demonstrating the operation | movement in the 1st assistance timing of a determination output part same as the above. It is a figure which shows the other example of specific selection operation same as the above. It is explanatory drawing for demonstrating the operation | movement in the 2nd assistance timing of a determination output part same as the above. It is a flowchart which shows operation | movement of the operation assistance determination part same as the above. It is a flowchart which shows operation | movement of the objective estimation apparatus same as the above. It is a block diagram which shows the structure of the conventional objective estimation apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100,1900 Purpose estimation apparatus 101 Input part 102 Choice input part 103 Output part 104 Choice presentation part 105 Operation support presentation part 106 Support request input part 107 Input / output control part 108 Operation history storage part 109 Equipment control part 110 Subgoal estimation part 111 Function Expression storage unit 112 Word concept storage unit 113, 1901 Purpose estimation unit 114 Operation support determination unit 201 Input interpretation unit 202 Neighboring element identification unit 203 Derivation unit 204 Estimation unit 301 Possibility base estimation unit 302 Certainty factor base estimation unit 303 Switching unit 502 Cursor key 503 Determination button 504 Back button 505 Power button 506 Menu button 507 Help button 601 Function expression database 1902 Operation support determination unit 1903 Time information acquisition unit

Claims (12)

Option acquisition means for acquiring, as an input option, an option selected by the user from among a plurality of options;
For each function of the operation target device, a functional element storage unit that stores a plurality of elements included in the character string of the function name of the function as an element group;
For each function of the operation target device, between each of a plurality of partial ranges composed of one or a plurality of the elements, and an input option acquired by the option acquisition unit, from the function element group A subgoal specifying means for specifying a subrange corresponding to the maximum similarity as a subgoal by calculating the similarity;
An object estimation device comprising: an estimation unit configured to estimate that a function having a higher degree of similarity of the subgoal among a plurality of functions of the operation target device is intended for a user. The subgoal specifying means includes:
By dividing the character string representing the input option acquired by the option acquisition unit into a plurality of elements, the plurality of elements are generated as input interpretations, respectively.
For each function of the operation target device, a similarity between each of the plurality of input interpretations of the input options and each of the plurality of partial ranges is calculated, and a partial range corresponding to the maximum similarity is calculated as the subgoal. The object estimation device according to claim 1, characterized by: The subgoal specifying means includes:
For each input option acquired by the option acquisition means, specify a subgoal of each function of the operation target device for the input option,
The estimation means includes
For each function of the operation target device, a maximum similarity among a plurality of subgoal similarities specified between the function and each input option is calculated as a probability score of the function, and the operation The objective estimation apparatus according to claim 2, further comprising possibility-based estimation means for estimating that a function having a higher possibility score among a plurality of functions of the target device is intended by the user. The subgoal specifying means includes:
For each input option acquired by the option acquisition means, specify a subgoal of each function of the operation target device for the input option,
The estimation means includes
For each function of the operation target device, the ratio of the number of elements that are not included in any of the plurality of subgoals specified between the function and each input option in the function element group is large. A certainty factor that is smaller and larger as the similarity between the plurality of subgoals is larger is calculated, and among the functions of the operation target device, it is estimated that the function having the larger certainty factor is the user's purpose. The objective estimation apparatus according to claim 2, further comprising a certainty factor base estimation unit. The purpose estimation device further includes:
Provided with a support request receiving means for receiving a support request from a user;
The subgoal specifying means includes:
For each input option acquired by the option acquisition means until the support request is received, identify a subgoal of each function of the operation target device for the input option,
The possibility-based estimation unit or the certainty factor-based estimation unit of the estimation unit includes:
When the request for assistance is received, the possibility score or the certainty factor is calculated, and it is estimated that the function having the larger possibility score or the certainty factor is more aimed at the user. The purpose estimation apparatus according to claim 3 or 4. The purpose estimation device further includes:
Providing means for presenting the function estimated by the estimating means;
The estimating means further includes:
For each function of the operation target device, a maximum similarity among a plurality of subgoal similarities specified between the function and each input option is calculated as a probability score of the function, and the operation Among a plurality of functions of the target device, a possibility-based estimation unit that estimates that a function having a higher possibility score is intended for the user;
When the certainty factor is larger than the first threshold value, the presenting unit is caused to present the function estimated by the certainty factor base estimating unit, and when the certainty factor is smaller than the first threshold value, the possibility base The purpose estimation apparatus according to claim 4, further comprising switching means for causing the presenting means to present the function estimated by the estimating means. The certainty factor base estimation unit updates the certainty factor of each function each time input options are sequentially acquired by the option acquisition unit,
The purpose estimation device further includes:
Presenting means for presenting the function estimated by the estimating means;
When the overall change amount of the certainty factor of each function updated by the certainty factor base estimation unit is smaller than a second threshold value, the function estimated by the certainty factor base estimation unit is presented to the presenting unit. The purpose estimation apparatus according to claim 4, further comprising: a presentation timing determination unit that performs the operation. The presentation timing determination means includes
When the change amount is smaller than the second threshold value and there is a certainty factor that is greater than a third threshold value among the certainty factors of the respective functions, the function estimated by the certainty factor base estimation unit is Let the presentation means present
The purpose estimation apparatus according to claim 7, wherein: The possibility-based estimation means updates the possibility score of each function each time input options are sequentially acquired by the option acquisition means,
The purpose estimation device further includes:
A time measuring unit that measures an elapsed time from the time when the option acquisition unit last acquired the input option when the input option is sequentially acquired by the option acquisition unit;
Presenting means for presenting the function estimated by the estimating means;
When the elapsed time measured by the time measuring unit is longer than a fourth threshold and the possibility score of each function is greater than a fifth threshold, the possibility-based estimation The purpose estimation apparatus according to claim 3, further comprising a presentation timing determination unit that causes the presentation unit to present the function estimated by the unit. The purpose estimation device further includes:
When there are a plurality of function names having a common subgoal among the function names of a plurality of functions estimated by the possibility-based estimation means, a plurality of function names having the common subgoal are collectively presented. The purpose estimation apparatus according to claim 3, further comprising a presentation unit. An option acquisition step in which the option acquisition unit acquires an option selected by the user from among a plurality of options as an input option;
The subgoal specifying unit, for each function of the operation target device, from a plurality of elements included in a character string of a function name of the function, each of a plurality of partial ranges including one or a plurality of the elements, and the option acquisition step A subgoal specifying step for specifying a subrange corresponding to the maximum similarity as a subgoal by calculating a similarity between the input option acquired in
And a presuming step in which an estimation unit estimates that a function having a higher degree of similarity of the subgoal among a plurality of functions of the operation target device is a user's purpose. An option acquisition step of acquiring an option selected by the user from among a plurality of options as an input option;
For each function of the operation target device, from the plurality of elements included in the character string of the function name of the function, each of a plurality of partial ranges including one or a plurality of the elements and the input acquired in the option acquisition step A subgoal specifying step for specifying a subrange corresponding to the maximum similarity as a subgoal by calculating a similarity between the options,
A program that causes a computer to execute an estimation step of estimating that a function having a higher degree of similarity of the subgoal among a plurality of functions of the operation target device is intended by a user.

Images