JP2007004394A - Product design support apparatus, product design support method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To support product design in a planning stage of a new product so that it will meet specifications close to a concept the new product should have. <P>SOLUTION: A relational modeling processing part 10 formulates a relational model function showing a causal relation for respective pieces of information by using the respective pieces of information on a concept keyword of a sold vehicle extracted from new product news, an accordance rate for a previously selected image word for a purchased vehicle acquired from a purchaser questionnaire, a form factor of the sold vehicle, and the sales share. A sales share prediction processing part 24 substitutes an input concept keyword for a new model vehicle into the relational model function for the image word to extract an image word for the new model vehicle, and substitutes an image word selected from them into the relational model function for the sales share to obtain predicted sales share. A vehicle shape specific processing part 22 substitutes the selected image word into the relational model function for the form factor to extract a form factor having large contribution to the image word, forms and displays an image of the vehicle shape, and further modifies the vehicle shape according to user's operation. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a product design support apparatus, and more particularly to support for determining product specifications.

  Manufacturers can generate profits when products are purchased by consumers. Therefore, in order for a product to be selected and purchased by a consumer, it is desirable to accurately grasp the needs from the consumer's preferences and the like and develop a product with specifications that meet the needs. For example, when an automaker intends to introduce a new model to the market, the new model's shape, etc. can be analyzed by analyzing the sales performance of existing cars and current consumer preferences at the planning stage of the new model. Tentatively determine the specifications of sales and predict sales share. When the expected sales share is not expected, the vehicle specifications are reviewed, and the sales share of the new model after the review is predicted again. In this way, the specifications of the new model are determined while obtaining the predicted sales share.

  For example, Patent Literature 1 discloses a technology for predicting a sales share in a product planning stage and examining vehicle specifications. In this patent document 1, the sales share prediction of a new model vehicle is accurately performed while associating the specifications, sales results, etc. of the already-sold vehicle with evaluations by consumers, and the specification of the new model vehicle is determined.

  By the way, in the planning stage of new vehicles, the vehicle shape is designed and details of specifications are determined with reference to information on already-sold cars and consumer evaluation, but the initial stage of vehicle planning. So, rather than suddenly working on the design of the vehicle shape, first of all, it is an abstract such as what kind of concept, what consumer group is targeted, and what kind of image vehicle to promote sales, It will be decided from conceptual matters. Therefore, when determining the specifications of a vehicle that predicts sales share, it is more preferable to consider abstract and conceptual matters in addition to information on already-sold vehicles, particularly competitive vehicles, and consumer evaluation. It is believed that there is.

Japanese Patent Application Laid-Open No. 2004-206569

  However, in the above-described conventional example, it is possible to determine the vehicle specification accepted by the consumer by considering the evaluation by the consumer, etc., but the information related to the abstract and conceptual vehicle concept and the vehicle image is not included. Since it was not incorporated as information when determining the specifications, it could not be said that the determined vehicle specifications matched the concept and image of the new vehicle.

  The present invention has been made to solve the above-mentioned problems, and its purpose is a product that supports product design so that the specifications close to the concept that the new product should have in the planning stage of the new product. It is to provide a design support apparatus.

  In order to achieve the above object, the product design support apparatus according to the present invention includes a concept information storage unit that stores, for each existing product, conceptual information that conceptually represents the characteristics of each existing product, and each existing product. Product information storage means for storing product information including specification information on at least the shape of each existing product, and the information stored in each of the storage means is substituted into a model function representing a causal relationship, and learned by conceptual information. Modeling means for formulating a relational model function indicating the causal relationship between the product information and product information, and substituting the input conceptual information of the new product into a relational model function indicating the causal relation between the conceptual information and the specification information. And design processing means for specifying and outputting the shape of the new product.

  The product information stored in the product information storage means further includes performance information related to the sales performance of each existing product, and the concept information of the input new product is converted into the causal result of the concept information and the performance information. It is characterized by having a prediction processing means for predicting sales of a new product by substituting it into a relational model function indicating the relation.

  Moreover, it has the concept information generation means which produces | generates the concept information stored in the said concept information storage means.

  Further, the concept information generation means extracts a keyword that can be a concept of each existing product from an information source specified in advance, and registers it in the concept information storage means as concept information of the existing product.

  Further, the concept information generation means calculates image evaluation data of the existing product from the evaluation result of the image of the existing product by the product purchaser, and registers it in the concept information storage means as concept information of the existing product. Features.

  The product design support apparatus according to the present invention is implemented by a computer used for product design, and includes conceptual information of each existing product that conceptually represents the characteristics of each existing product and specification information about at least the shape of each existing product. An existing product information input step for inputting product information of each existing product, and a relationship indicating the causal relationship between conceptual information and product information by substituting each input information into a model function that represents the causal relationship and learning. A modeling step for formulating a model function, a new product information inputting step for inputting conceptual information of a new product, and conceptual information and specification information formulated by inputting the conceptual information of the new product by the modeling step. And a design step for specifying and outputting the shape of the new product by substituting it into a relational model function indicating the causal relationship with

  According to the present invention, the shape of the new product is designed by incorporating the conceptual information about the concept of the new product and the image to be aimed at, so that the shape of the new product that matches the concept of the new product can be obtained. it can. In addition, it is possible to improve the accuracy of sales prediction performed in the process of obtaining the shape of the new product. As a result, the quality of new product planning can be improved and the efficiency of planning work can be improved.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In the present embodiment, a case will be described as an example where the present invention is applied to vehicle shape determination support in the planning stage of a new vehicle, which is a new product, particularly in the design stage of a new vehicle.

  FIG. 1 is a block diagram showing a product design support apparatus according to the present embodiment. In the present embodiment, a modeling process for formulating a relational model function indicating the causal relationship of information, and a shape of the new vehicle is determined while performing a simulation using the relational model function modeled by the modeling process. It can be roughly divided into design processing. In addition, “decision” here refers to narrowing down the shape of the new model to only one as a result of the processing described later by the product design support device, which is the final result of work in the planning stage. It doesn't mean. Further, main information handled in the present embodiment includes conceptual information and vehicle information. As will be described in detail later, the concept and image of the vehicle correspond to the concept information. The vehicle information is information on each sold vehicle and includes specification information and performance information. The specification information is information related to the shape and specifications of each vehicle, but in this embodiment, the form information calculated using the specification information and the specification information is also handled as one element of the specification information. The track record information is information obtained when sold vehicles such as the number of vehicles sold, and the sales share is used in the present embodiment.

  Among the above processes, the modeling process includes a concept keyword extraction processing unit 2, a vehicle image evaluation aggregation processing unit 4, a form element calculation processing unit 6, a sales share calculation processing unit 8, a relational modeling processing unit 10, and a relational model correction processing unit. 12 is implemented. On the other hand, the design process is performed by the design processing unit 20 including the vehicle shape identification processing unit 22 and the sales share prediction processing unit 24.

  The concept keyword extraction processing unit 2 extracts the new product news issued when the existing vehicle is released from the new product news storage unit 30, extracts the keywords corresponding to the terms indicating the concept of each existing vehicle, Each vehicle is registered in the concept information storage unit 32. A competing vehicle of the new vehicle specified by determining the concept of the new vehicle corresponds to the “existing vehicle” handled in the present embodiment. The vehicle image evaluation aggregation processing unit 4 evaluates the image of each existing vehicle by aggregating the questionnaire results to the purchaser of each existing vehicle stored in the questionnaire result storage unit 34, and stores the evaluation result as image information. Register in the part 36. The form element calculation processing unit 6 registers the specification information of each existing vehicle extracted from the vehicle database 38 as a morpheme element and, if necessary, the form element obtained by calculation from the specification information in the form element information storage unit 40. The sales share calculation processing unit 8 calculates the sales share of each existing vehicle from the sales performance information of each existing vehicle extracted from the vehicle database 38 and registers the calculation result in the sales share information storage unit 42. The relationship modeling processing unit 10 substitutes various types of information obtained by the above processing units 2 to 8 into model functions representing causal relationships and learns them by using concept keywords, image words, and specification information that are conceptual information. A relational model function indicating a causal relationship between a certain form element and sales share as performance information is formulated. Then, each formulated relational model function is registered in the modeling information storage unit 44. The relationship model function related to the image word among those formulated by the relationship modeling processing unit 10 is based on a questionnaire result at the time of selling each existing vehicle, and there may be a deviation from the current image sense. obtain. Therefore, the relational model correction processing unit 12 corrects the relational model function related to the image word registered in the modeling information storage unit 44 based on the questionnaire result stored in the correction questionnaire result storage unit 46. The questionnaire whose result is stored in the correction questionnaire result storage unit 46 generates a virtual vehicle based on the design information (CAD data) of the existing vehicle stored in the vehicle database 38, and in the near future, relates to an image of the virtual vehicle. This is an implemented questionnaire.

  The design processing unit 20 obtains processing results such as the shape of the new model vehicle and the predicted sales share by using the relational model function obtained as a result of the processing by the processing units 2 to 10 described above. The vehicle shape specification processing unit 22 included in the design processing unit 20 changes the CAD data of the vehicle according to a user operation, and finally outputs the shape of the new model vehicle. The sales share prediction processing unit 24 calculates the sales share of the new vehicle.

  The product design support apparatus in the present embodiment is realized by a computer having a CAD function such as a CAD terminal apparatus. The processing functions in the processing units 2 to 12 and 20 are realized by a cooperative operation between a computer and a program executed by the CPU of the computer. Each storage unit 30 to 44 including the vehicle database is realized by an external storage device such as an HDD, but does not need to be directly connected to the product design support device, and can be accessed via a network when necessary. If it is.

  What is characteristic in this embodiment is not only the sales results of competing vehicles when determining the specifications of the new vehicle, especially the vehicle shape, and the concept and image of the new vehicle, but also the questionnaire to the vehicle purchaser. By taking into account this information, it is possible to identify the shape of a new vehicle that can secure the expected sales share.

  Hereinafter, the operation of this characteristic embodiment will be described. In the present embodiment, as described above, the modeling process for generating the relational model function used when determining the shape of the new model vehicle and the shape of the new model vehicle using the relational model function modeled by this modeling process are used. Since it can be roughly divided into the design process to be determined, the modeling process will be described first with reference to the flowchart shown in FIG.

  In step 110, the concept keyword extraction processing unit 2 extracts a keyword indicating the concept of each existing vehicle from the information related to each existing vehicle and registers it in the concept information storage unit 32 for each existing vehicle. Process as follows.

  First, in the present embodiment, new product news is used as an information source from which keywords are extracted. This is because the new product news is an article about the existing vehicle that was announced when each existing vehicle was introduced to the market, and features, specifications, and concepts of the existing vehicle were introduced. Therefore, the concept keyword extraction processing unit 2 extracts the new product news of the existing vehicle from the new product news storage unit 30, extracts keywords corresponding to terms indicating the concept of each existing vehicle, and registers them in the concept information storage unit 32. To do. This extraction and registration process is performed for each existing vehicle.

  FIG. 3 is a diagram showing an example of the content of concept information stored in the concept information storage unit 32. The concept information is generated by associating T / C and one or more concept keywords extracted by the concept keyword extraction processing unit 2 with a vehicle name for identifying each existing vehicle as shown in FIG. Note that T / C (Time Control) in FIG. 3 indicates the sales period of the vehicle, and the start time of the vehicle and the end time if sales have ended are set. The same applies to T / C in other diagrams showing information setting contents.

  As a keyword extraction method in the concept keyword extraction processing unit 2, there is a method of automatically detecting nouns and adjectives from a text sentence related to a product concept and then automatically excluding duplicated terms. Of course, the result of the automatic extraction processing by the concept keyword extraction processing unit 2 may be used as a concept keyword as it is, but it is left to human judgment whether the extracted keyword correctly represents the concept of the vehicle. Also good. In other words, for example, the keywords extracted by the concept keyword extraction processing unit 2 are displayed in a list on a display (not shown), and a new vehicle planner (hereinafter simply “user”) is selected from the list as an appropriate keyword as a concept. You may comprise as follows. By allowing the user to finally determine the terms to be extracted, the validity, accuracy, accuracy, etc. of the concept keywords can be improved.

  In the present embodiment, new product news is used as a predetermined information source. However, the information source need not be limited to new product news. Further, in the present embodiment, it has been described that information is extracted from the storage unit inside the apparatus. However, for example, information may be extracted from other information sources such as acquiring information via the Internet.

  In step 120, the vehicle image evaluation aggregation processing unit 4 evaluates the image of each existing vehicle and registers the evaluation result in the image information storage unit 36. Specifically, the vehicle image evaluation totalization processing unit 4 performs the following processing.

  Car dealers conduct questionnaires on consumers who have purchased existing vehicles to evaluate the degree of match of purchased vehicles with “image words”. The results of this questionnaire are digitized and accumulated in the questionnaire result storage unit 34. “Image word” is, for example, “sporty”, “sophisticated”, or a term that expresses an image of each vehicle. These terms are selected in advance in the questionnaire as evaluation items. include. The purchaser then scores an image of the purchased vehicle. In the present embodiment, each image word is given a high evaluation value so as to match the image word with a maximum score of 10 points.

  The vehicle image evaluation aggregation processing unit 4 extracts these questionnaire results from the questionnaire result storage unit 34, aggregates them for each existing vehicle image word, obtains an average value, and calculates the average value obtained for the vehicle corresponding to the “image word”. And is registered in the image information storage unit 36.

  FIG. 4 is a diagram showing an example of the contents of the image information stored in the image information storage unit 36. The image information includes, as shown in FIG. 4, the T / C and the matching degree (average value of evaluation values) for each image word calculated by the vehicle image evaluation aggregation processing unit 4 in the vehicle name for identifying each existing vehicle. Generated in association.

  In step 130, the form element calculation processing unit 6 obtains the form element of each existing vehicle and registers it in the form element information storage unit 40. Specifically, the form element calculation processing unit 6 performs the following process.

  First, the “form element” in the present embodiment refers to a specific dimension value or a combination of dimension values when designing a vehicle. In the present embodiment, since the vehicle shape is displayed as an image using the form element value, information items necessary for displaying the vehicle shape image are set in advance as at least form element data items.

  FIG. 5 is a diagram showing an example of the contents of the form element information registered in the form element information storage unit 40. As shown in FIG. 5, the morphological elements are calculated as “total height / full length”, for example. Since the original data for obtaining these form elements is included in the specification information, the form element calculation processing unit 6 extracts the specification information of each vehicle from the vehicle database 38, and obtains preset form elements by calculation. Alternatively, the extracted specification information is used as a form element as it is. The morphological element calculation processing unit 6 registers the morphological element thus obtained in the morphological element information storage unit 40. In the present embodiment, each morphological element value is normalized to a maximum of 10 points for subsequent modeling. As shown in FIG. 5, the morpheme element information is generated by associating the T / C and the value of each morphological element obtained by the morphological element calculation processing unit 6 with the vehicle name for identifying each existing vehicle.

  In step 140, the sales share calculation processing unit 8 calculates the sales share and the like of each existing vehicle, and registers the calculation result in the sales share information storage unit 42. Specifically, the processing is performed as follows.

  That is, the sales share calculation processing unit 8 extracts the sales performance information of each existing vehicle from the vehicle database 38, and calculates the monthly sales share (unit%) from the release month. As described above, an existing vehicle is a competitor of a new vehicle and belongs to the same category (3 Box, large, etc.) as the new vehicle, so the denominator when calculating the sales share is all Sales volume and numerator are the sales volume of the vehicle. And the sales share calculation process part 8 calculates the average value of a sales share by dividing the sum total of the share value of each month by the number of months. The sales share calculation processing unit 8 registers the sales share and average value of each month obtained in this way in the sales share information storage unit 42.

  FIG. 6 is a diagram showing an example of the contents of the sales share information registered in the sales share information storage unit 42. In the sales share information, as shown in FIG. 6, the T / C, the average value of the sales share, and the monthly sales share obtained by the sales share calculation processing unit 8 are associated with the vehicle name for identifying each existing vehicle. Generated.

  In addition, since each said step 110-140 can be implemented independently, it does not necessarily need to process in the order shown in FIG. 2, and may process it in parallel.

  In step 150, the relational modeling processing unit 10 formulates a relational model function using the various information obtained in the above steps 110 to 140 and registers it in the modeling information storage unit 44. Process like this.

  FIG. 7 is a diagram illustrating an example of each relational model function registered in the modeling information storage unit 44. In this embodiment, since each value of concept keyword, image word, form element, and sales share is obtained by the above steps 110 to 140, each value is input to the corresponding relational expression shown in FIG. Formulate by analysis or learning by neural network. In the present embodiment, a relational model function (FIG. 7 (1)) showing a causal relationship between a concept keyword and an image word, a relational model function (FIG. 7 (2)) showing a causal relation between an image word and a form element, A relational model function (FIG. 7 (3)) indicating the causal relationship between the form element and the sales share and a relational model function (FIG. 7 (4)) indicating the causal relation between the image word and the sales share are respectively formulated.

  In step 160, the relational model correction processing unit 12 corrects the relational model function related to the image word registered in the modeling information storage unit 44. The reason for correcting the relational model function is that, as described above, it is considered that the sense of a human image or the like changes with the passage of time. For example, the shape of a vehicle that is felt sporty at the time of selling an existing vehicle does not always feel sporty without changing. Therefore, if the evaluation at the time of sale of the existing vehicle is used as it is, there is a possibility that a sense shift occurs in the finally obtained output. Thus, in the present embodiment, the relational model correction processing unit 12 is provided to correct the relational model function in order to correct the shift in feeling with the passage of time. Specifically, the correction of the relational model function is processed as follows.

  First, the relational model correction processing unit 12 specifies one existing vehicle as a reference vehicle by automatic selection or user designation, and changes one or more virtual element data values by changing one or more form element data values of the reference vehicle. A vehicle is generated and the vehicle shape is displayed on the screen. Since the CAD data for displaying the virtual vehicle shape can be obtained from the vehicle database 38 as design information (CAD data) of the reference vehicle, it can be generated by changing this data. Then, the questionnaire target person is asked to input the evaluation value for each image word as it feels when viewing the virtual vehicle. Since there is no purchaser of a virtual vehicle, for example, a questionnaire is conducted with employees of a manufacturer as a questionnaire target person. The questionnaire results are accumulated in the correction questionnaire result storage unit 46. The relational model correction processing unit 12 corrects the relational model function related to the image word registered in the modeling information storage unit 44 based on the questionnaire result stored in the correction questionnaire result storage unit 46. That is, similar to step 150, the form factor value of each virtual vehicle and the evaluation value for each image word, which are the questionnaire results, are substituted into the relational model function (FIG. 7 (2)) relating to the image word to be learned. In this way, the accuracy of the relational model function is improved by bringing the evaluation with the sense obtained by the past questionnaire closer to the evaluation with the current sense. Thereby, the relational model function is updated so as to obtain an output that matches the current sense.

  In the present embodiment, for convenience, the model function correction processing in the relational model correction processing unit 12 has been described with reference to FIG. 2, but the processing in the relational model correction processing unit 12 is for improving the accuracy of the model function. Therefore, it is not always necessary to carry out as a series of processes following step 150. Alternatively, it may not be carried out or may be carried out in advance independently of the above processes.

  In the relational modeling process of the present embodiment, the relational modeling function is generated as described above. Next, a design process for a new vehicle in the present embodiment using the generated relational modeling function will be described with reference to the flowchart shown in FIG. The design process in the present embodiment includes a process for setting a predicted sales share of a new vehicle by the sales share prediction processing unit 24, and a process for specifying the shape of a new vehicle that achieves the predicted sales share by the vehicle shape specifying processing unit 22. Divided into

  In step 210, the sales share prediction processing unit 24 accepts a concept keyword input by the user using a user interface (not shown) such as a mouse, a keyboard, or a display. The inputted concept keywords are one or more terms expressing the concept of a new type vehicle devised by the user.

  In step 220, the sales share prediction processing unit 24 substitutes the concept keyword of the new model vehicle into the equation (1) shown in FIG. 7 among the information function model functions registered in the modeling information storage unit 44. Extract image words that greatly contribute to the concept. Since the output value increases as the contribution increases, the image words are arranged and displayed on the screen in the descending order of the output value, that is, in the descending order of the contribution to the concept of the new vehicle. The image word to be displayed is assumed to have an output value equal to or greater than a preset threshold value. The image word to be displayed is temporarily stored for use in later processing.

  In step 230, the sales share prediction processing unit 24 calculates the sales share based on the image word selected by the user. That is, when the user selects an image word that matches the concept of the new vehicle from among the image words displayed in the list, the sales share prediction processing unit 24 selects the selected image word using the equation (4) shown in FIG. When the sales share is obtained by substituting in (), this sales share is displayed on the screen.

  In step 240, the sales share prediction processing unit 24 inquires of the user about the validity of the predicted value of the sales share together with the display of the sales share. If the user wants to execute the above process (steps 210 to 230) again because the predicted value of the sales share is not more than expected, the user performs an input operation to that effect and obtains a sales share higher than expected. Therefore, if it is desired to proceed to the vehicle design, an input operation to that effect is performed. The sales share prediction processing unit 24 shifts the processing to step 210 or step 250 depending on the contents of this input operation.

  In step 250, the vehicle shape identification processing unit 22 converts the image word selected by the user in step 230 into equation (2) shown in FIG. 7 among the information function model functions registered in the modeling information storage unit 44. By substituting, a morphological element value having a large contribution to the image word is extracted.

  In step 260, the vehicle shape identification processing unit 22 forms a vehicle shape image according to the form element value extracted in step 250 and displays it on the screen.

  In step 270, the vehicle shape specification processing unit 22 inquires of the user about the validity of the vehicle shape designed together with the display of the vehicle shape image. In response to this inquiry, the user inputs the appropriateness of the vehicle shape displayed on the screen. If it is input that the displayed vehicle shape is acceptable, the vehicle shape identification processing unit 22 ends the process after outputting predetermined information. The output information will be described later.

  On the other hand, if it is input that the displayed vehicle shape is “non-”, in step 280, whether the vehicle shape identification processing unit 22 changes the morphological element value or the image word to change the vehicle shape. Inquire about.

  Here, when it is input that the form element value is to be changed, in step 290, the vehicle shape identification processing unit 22 causes the user to change the form element value. As a method for changing the form element value, the numerical value data of the form element extracted in step 290 may be changed by keyboard input or the like. In the present embodiment, the form element value is changed according to the user operation using the mouse. The shape of the vehicle image displayed on the screen is changed. For example, the user changes the shape such as increasing the vehicle height by dragging and dropping. In this way, the user can easily change the vehicle shape. Since the vehicle shape on the screen can be grasped as CAD data, the form element value can be calculated from the dimension value on the screen. Thereafter, in step 260, a vehicle shape image is formed according to the form element value changed as described above, and is displayed on the screen. In the present embodiment, since the shape of the vehicle on the screen is changed, the changed shape is already displayed on the screen.

  On the other hand, when it is input that the image word is to be changed, in step 300, the vehicle shape identification processing unit 22 displays a list of image words to be displayed in step 220 on the screen and causes the user to reselect the image word. In step 230, since the predicted sales share is obtained using the selected image word, the predicted sales share value may be greatly changed if the image word to be reselected is significantly switched. In such a case, the vehicle design process may be restarted from the beginning. Thereafter, in step 250, morphological element values having a large contribution to the image word reselected by the user as described above are extracted.

  When the user can design a satisfactory vehicle shape, in step 270, the user inputs that the vehicle shape displayed in response to the inquiry from the vehicle shape identification processing unit 22 may be used.

  When the shape of the new model vehicle is specified as described above, in step 310, the vehicle shape specification processing unit 22 selects the concept keyword, the selected image word, the predicted sales share, and the final value input in the above processing. The design information (CAD data) indicating the vehicle shape obtained in advance is output to a previously designated destination such as a printer or an external storage device, and the process is terminated.

  According to the present embodiment, as described above, conceptual information about the concept of the new vehicle and the image to be aimed at is taken in as input information when forecasting the sales share and designing the new vehicle. The shape of the new model that matches the image can be obtained. Further, it is possible to improve the accuracy of the predicted sales share calculated in the process of obtaining the vehicle shape. As a result, the quality of the new vehicle product planning can be improved and the efficiency of the planning work can be improved.

  In this embodiment, the case where the product design support device is used for the design support of the new vehicle in the planning stage of the new vehicle is taken as an example. However, the present invention can be used in other stages as well. It can also be applied to other product development at other manufacturers.

It is the block block diagram which showed one Embodiment of the product design support apparatus which concerns on this invention. It is the flowchart which showed the modeling process in this Embodiment. It is the figure which showed the example of the content of the concept information in this Embodiment. It is the figure which showed the example of the content of the image information in this Embodiment. It is the figure which showed the example of the content of the form element information in this Embodiment. It is the figure which showed the example of the content of the sales share information in this Embodiment. It is the figure which showed the example of the relationship modeling function in this Embodiment. It is the flowchart which showed the new vehicle design process in this Embodiment.

Explanation of symbols

  2 concept keyword extraction processing unit, 4 vehicle image evaluation aggregation processing unit, 6 form element calculation processing unit, 8 sales share calculation processing unit, 10 relationship modeling processing unit, 12 relationship model correction processing unit, 20 design processing unit, 22 vehicle Shape identification processing unit, 24 sales share prediction processing unit, 30 new product news storage unit, 32 concept information storage unit, 34 questionnaire result storage unit, 36 image information storage unit, 38 vehicle database, 40 form element information storage unit, 42 sales Share information storage unit, 44 Modeling information storage unit, 46 Correction questionnaire result storage unit.

Claims (6)

Concept information storage means for storing concept information that conceptually represents the characteristics of each existing product for each existing product;
Product information storage means for storing, for each existing product, product information including specification information on at least the shape of each existing product;
Modeling means for formulating a relational model function indicating the causal relation between the concept information and the product information by substituting the information stored in each storage means into a model function representing the causal relation, and learning;
Design processing means for identifying and outputting the shape of the new product by substituting the input conceptual information of the new product into a relational model function indicating the causal relationship between the conceptual information and the specification information,
A product design support apparatus characterized by comprising: The product design support apparatus according to claim 1,
The product information stored in the product information storage means further includes performance information related to the sales performance of each existing product,
Product design support characterized by having prediction processing means for predicting sales of new products by substituting the input concept information of new products into a relational model function that indicates the causal relationship between concept information and performance information apparatus. The product design support apparatus according to claim 1,
A product design support apparatus comprising concept information generation means for generating concept information to be stored in the concept information storage means. The product design support apparatus according to claim 1,
The concept information generation means extracts a keyword that can be a concept of each existing product from an information source specified in advance, and registers it in the concept information storage means as concept information of the existing product. . The product design support apparatus according to claim 1,
The concept information generation means calculates image evaluation data of the existing product from an image evaluation result for the existing product by a product purchaser, and registers it in the concept information storage means as concept information of the existing product. Product design support device. Implemented by a computer used for product design,
An existing product information input step for inputting product information of each existing product including conceptual information of each existing product conceptually expressing the characteristics of each existing product and specification information regarding at least the shape of each existing product;
A modeling step that formulates a relational model function that indicates the causal relationship between conceptual information and product information by substituting each input information into a model function that represents the causal relationship, and learning,
A new product information input step for inputting concept information of the new product;
Design that identifies and outputs the shape of the new product by substituting the input conceptual information of the new product into the relational model function that indicates the causal relationship between the conceptual information formulated in the modeling step and the specification information. Steps,
A product design support method comprising:

Images