JP2013101226A - Real estate property information provision support device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow more accurate search of real estate property information to be provided to a guest on the basis of voices of the guest and a host.SOLUTION: An H/G determination part 41 identifies voices of a host and a guest who are seated at different positions on the basis of intensity of detection sounds by directional microphones 22 to 25 whose sound collecting directions are freely adjustable. Question keywords and answer keywords of each of the types of questions classified with one or more question keywords are registered in a keyword table 45. A keyword determination part 44 determines which of the registered question keywords is included in the identified voice of the host and the type of the question based on the question keyword, determines which of the registered answer keywords corresponding to the type of the latest question is included in the identified voice of the guest, and makes a keyword for search storage part 46 store the determined answer keyword in association with the type of the question.

Description

  The present invention relates to a real estate property information providing support apparatus that automatically searches real estate property information based on keywords included in a voice.

  In Patent Document 1 below, a microphone is placed on the customer side who wants to purchase real estate, a salesperson interviews, extracts keywords from voice information issued by the customer, and based on this, a real estate property information database The real estate property information provision apparatus which searches for and presents the corresponding information is disclosed.

  According to this apparatus, since it is not necessary for a sales person to input a keyword to the real estate property information search device with a keyboard or the like, real estate property information can be provided without hindering communication with the customer. .

JP-A-2005-321916

  However, since it is not distinguished whether the keyword is issued by a customer or a sales person, the search result becomes inaccurate.

  In view of such problems, an object of the present invention is to provide a real estate property information providing support device capable of more accurately searching for real estate property information to be provided to a guest based on the voice of the guest and the host. There is.

In one embodiment of the present invention,
A keyword storage means for identification in which keywords are registered;
Keyword discriminating means for detecting voice and discriminating whether or not any of the registered keywords is included in the voice;
Search keyword storage means for storing the determined keyword;
A search start determination means;
An information database in which real estate property information is registered;
Search means for searching the information database for information satisfying a condition based on the keyword stored in the search keyword storage means in response to the search start determination means determining the search start;
Output means for outputting information search results in a predetermined format and medium;
In the real estate property information providing support device equipped with
A speaker identification means for identifying the voice of the host and the guest;
In the keyword storage means, a question keyword and an answer keyword for each question type classified by one or more question keywords are registered,
The keyword discriminating unit discriminates which of the registered question keywords is included in the voice of the identified host and the type of the question based on the question keyword. It is determined which of the registered answer keywords corresponding to the most recent question type is included, and the determined answer keyword is stored in the search keyword storage means in association with the question type.

  According to the configuration of the above aspect, the speaker identification means for identifying the voices of the host and the guest is provided, and the keyword storage means includes an answer keyword for each type of question classified by the question keyword and one or more question keywords. Is registered, and the keyword discriminating unit discriminates which of the registered question keywords is included in the voice of the identified host and the type of question based on the question keyword, and It is determined which of the registered answer keywords corresponding to the type of the most recent question is included in the voice, and the determined answer keyword is stored in the search keyword storage means in association with the type of question. Therefore, real estate property information can be searched more accurately, and there is an effect that the utility is high.

  Other objects, characteristic configurations and effects of the present invention will become apparent from the following description read in connection with the appended claims and the drawings.

1 is a schematic block diagram illustrating an arrangement of an image forming apparatus in an interview room according to Embodiment 1 of the present invention. It is a general | schematic functional block diagram of the real estate property information provision assistance apparatus in FIG. FIG. 3 is a schematic functional block diagram illustrating a configuration of a listening start / end determination / output start determination unit in FIG. 2. (A) and (B) are explanatory diagrams for determining whether or not the host faces the image forming apparatus. It is a functional block diagram which shows schematic structure of the keyword acquisition part in FIG. It is a schematic flowchart which shows the process by the keyword discrimination | determination part in FIG. FIG. 3 is a schematic functional block diagram illustrating a configuration of an information search processing unit in FIG. 2. It is a schematic flowchart which shows the process by the condition relaxation part in FIG. It is a schematic block diagram which shows the structural example of the process parts, such as a face information in FIG. (A)-(D) are explanatory drawings of the structure of the information database used in FIG. FIG. 3 is a schematic functional block diagram illustrating a configuration example of an information processing unit in FIG. 2. It is a schematic flowchart of the whole operation | movement of a real estate property information provision assistance apparatus.

  FIG. 1 is a schematic block diagram illustrating an arrangement of an image forming apparatus in an interview room according to the first embodiment of the present invention.

  In this interview room, the image forming apparatus 10 is arranged near the table 11, and the chair 11 for the host H and the chairs 13 and 14 for the guest G are combined on the table 11.

  The image forming apparatus 10 includes a main body 15 and a real estate property information providing support apparatus 20. This device 20 is configured by connecting an input device and an output device to an information processing device 21. This input device includes directional microphones 22 to 25 whose sound collection direction can be manually adjusted, and video cameras 26 and 27 whose optical axis direction can be manually adjusted.

  All of the directional microphones 22 to 25 are preferably unidirectional. The directional microphones 22 to 25 are arranged so that the azimuths at which the respective sensitivities are maximized are radial, and based on the output intensity of the directional microphones 22 to 25 without determining the tone of the host H. The voice of the host H sitting on the chair 12 and the voice of the guest G sitting on the chair 13 or 14 can be distinguished.

  The optical axis of the video camera 26 is directed to the chair 12 side, and it can be determined by image processing whether the host H is seated on the chair 12 and whether the host H is facing the image forming apparatus 10 side. It has become. The video camera 27 has its optical axis directed toward the chairs 13 and 14, and determines whether or not the guest G is seated on the chairs 13 and 14 and the image of the guest G will be described later. It is possible to process personal identification and gender estimation.

  The output device includes a speaker 28 for assisting questions and a display panel 29 for displaying the number of corresponding properties in the automatic search result. The display panel 29 is disposed on the image forming apparatus 10 and the image forming apparatus 10 is disposed on the table 11 side so that the host H sitting on the chair 12 can visually recognize the display on the display panel 29.

  FIG. 2 is a schematic functional block diagram of the real estate property information providing support device 20.

  The hearing start / end determination / output start determination unit 30 of the information processing apparatus 21 processes the image data from the video cameras 26 and 27 to determine the start and end of the listening and the image forming apparatus main body 15 or via e-mail. Judge the output start.

  FIG. 3 is a schematic functional block diagram showing the configuration of the listening start / end determination / output start determination unit 30.

  In the determination unit 30, the frame image in the temporary storage unit (frame memory) 31 is updated with the video data from the video cameras 26 and 27, while the chair image storage unit 32 stores the images of the chairs 12 to 14 in advance. Stored.

  The hearing start determination unit 33 compares the image in the temporary storage unit 31 with the chair image in the chair image storage unit 32 to identify the chairs 12 to 14 before the person is seated, and to the chairs 12 to 14. It is determined whether or not is seated. If the determination unit 33 determines that a person is seated on the chair 12 and that a person is seated on the chair 13 or 14, the determination unit 33 sets the listening start START (signal or variable) to “true” (active). Whether or not the user is seated is determined by whether or not a chair is determined to be a face and the height is within a predetermined range.

  The output start determination unit 34 extracts the face image of the host H sitting on the chair 12 from the frame image from the video camera 26 included in the temporary storage unit 31, and the host H keeps the image forming apparatus 10 side for a predetermined time. For example, it is determined whether or not the camera is continuously turned for 3 seconds or more, and when an affirmative determination is made, the output start OUT (signal or variable) for outputting the search result is activated.

  Whether the host H is facing the image forming apparatus 10 is determined by identifying a pair of eyes included in the face image, for example, as shown in FIG. For example, it is determined that the host H is facing the image forming apparatus 10 side. If both ears are included in the face image, it may be determined that the image forming apparatus 10 is facing.

  When it is determined that no person is seated on any of the chairs 12 to 14 and this state has elapsed for a predetermined time, for example, 30 seconds, the end determination unit 35 activates the listening end END (signal or variable).

  Returning to FIG. 2, the outputs of the directional microphones 22 to 25 are respectively amplified by the amplifiers 221, 231, 241 and 251 having the same configuration, and further, A / D converters 222, 232 and 242 having the same configuration. And 252 are converted into digital values and supplied to the keyword acquisition unit 40 as audio data.

  Whether the keyword acquisition unit 40 is a question from the host H sitting on the chair 12 or an answer from the guest G sitting on the chair 13 or 14 based on the comparison of the amplitudes of these audio data If it is a question, the type of the question is determined. If the answer is within the set time after the question determination, an answer keyword related to the question type is determined, and the determined answer keyword is set as the question type. Store in association with each other.

  FIG. 5 is a functional block diagram illustrating a schematic configuration of the keyword acquisition unit 40.

  The H / G determination unit 41 compares the amplitudes of the audio data from the A / D converters 222, 232, 242 and 252 to determine whether the audio is the host H audio or the guest G audio. For example, the absolute value of the difference between the sum of the output amplitudes of the A / D converters 222 and 232 and the sum of the output amplitudes of the A / D converter 252 and the A / D converter 242 is greater than or equal to a set value. If it is positive, it is determined that the sound is from the guest G. If it is negative, it is determined that the sound is from the host H. If the difference between the output amplitudes of the A / D converters 232 and 242 is less than a certain value, it is determined as noise. Outputs from the A / D converters 222, 232, 242 and 252 are added by the adding unit 42 and stored in the temporary storage unit 43.

  Here, the property information to be searched includes a keyword, a floor plan, and a building exterior photograph. This keyword includes rent, nearest station, walking time from nearest station, floor plan, and building type.

  The types of questions from the host H to the guest G include the rent, the nearest station of the desired property, the walking time from the desired property to the nearest station, the floor plan of the desired property, and the type of building corresponding to the keyword. For these types, the host H asks the guest G the following questions.

Rent: “How much is the desired rent?”
Nearest station: “Is there any hope for the nearest station?”
Walking time: “How many minutes on foot are allowed from the nearest station?”
Floor plan: “How much would you like to plan?”
Building type: “Would you like an apartment, condominium, or detached house?”
When the H / G determination unit 41 determines that the sound is from the host H, the keyword determination unit 44 determines the type of the question based on the contents of the temporary storage unit 43 and the question keyword registered in the keyword table 45. To do.

  For example, if the host H's question contains “hope” and “rent” question keywords in the voice, it is determined that the question type is rent, and “hope” and “nearest station” question keywords are included. If the answer is yes, it is determined that the question type is the nearest station. If the question keyword “how many minutes on foot” is included, the question type is determined to be walking time. ”Question keyword, it is determined that the type of question is floor plan, and if any of“ Apartment ”,“ Apartment ”,“ Detached ”and“ hope ”question keywords are included, the question Is determined to be a building type. Further, as shown in FIG. 6, the keyword discriminating unit 44 sets a flag F of the discriminated question type and resets the flag F that was set last time.

  For example, the following answer keywords are registered in the keyword table 45 for each question type of rent, nearest station, walking time, floor plan, and building type.

Rent: "10,000", "20,000", ..., "100,000" (10 steps from 10,000 to 100,000, in units of 10,000 yen)
Nearest station: "Shinjuku", "Shibuya", ... (all stations on the Yamanote Line)
Walking time: "1 minute", "2 minutes", ..., "20 minutes" (20 steps from 1 to 20 minutes, 1 minute unit)
Floor plan: "1K", "1DK", "1LDK", "2K", "2DK", "2LDK", "3DK", "3LDK"
Building type: “apartment”, “apartment”, “detached”
When the keyword determination unit 44 determines that the H / G determination unit 41 is the voice of the guest G, the contents of the temporary storage unit 43 and those related to the latest question keyword among the answer keywords registered in the keyword table 45 Based on the above, the answer keyword is determined. The keyword determination unit 44 further determines whether or not the answer keyword (current answer keyword) is the same as the answer keyword (previous answer keyword) related to the question keyword stored in the keyword storage unit 46. If a negative determination is made, the previous keyword is overwritten and updated with the current answer keyword, and the search start SEARCH (variable or signal) is activated.

  For example, the initial values in the keyword storage unit 46 are as follows (assuming that the nearest station is the location of the real estate agent). These are included in the type of the most recent question keyword and the previous answer keyword.

Rent: Unconditional Nearest station: Shinjuku Walking: Unconditional Floor plan: Unconditional Building type: Unconditional FIG. 6 is a schematic flowchart showing processing by the keyword discriminating unit 44.

  (S0) If the listening end END from the listening start end determination / output start determining unit 30 is active, this process ends. Otherwise, the process proceeds to step S1.

  (S1) If the determination result of the H / G determination unit 41 is the host H, the process proceeds to step S2, and if it is the guest G, the process proceeds to step S4.

  (S2) It is determined whether or not the question keyword is included in the voice in the temporary storage unit 43. If it is included, the process proceeds to step S3, and if not, the process returns to step S0.

  (S3) It is determined as described above whether one or two question keywords are of the above-mentioned question type, and if any of them is met, the flag F is set and the flag F that was set last time is reset. Thus, the type of the current question is stored, and the process returns to step S0.

  (S4) It is determined whether or not the answer keyword regarding the current question type is included in the voice in the temporary storage unit 43. If included, the process proceeds to step S5, and if not, the process returns to step S0.

  (S5) The current answer keyword is compared with the corresponding previous answer keyword stored in the keyword storage unit 46. If they match, the process returns to step S0, and if not, the process proceeds to step S6.

  (S6) The answer keyword is updated by overwriting the corresponding previous answer keyword in the keyword storage unit 46 with the current answer keyword.

  (S7) Activate search start SEARCH and return to step S0.

  Returning to FIG. 2, the information search processing unit 50 AND-searches the information database with the keyword group in the keyword storage unit 46 in response to activation of the search start SEARCH, and outputs the result.

  FIG. 7 is a schematic functional block diagram showing the configuration of the information search processing unit 50.

  In response to the activation of the search start SEARCH, the information search unit 51 satisfies the predetermined condition of all keywords in the keyword storage unit 46 (for example, the condition of the keyword “3 minutes” is satisfied within 3 minutes on foot, (Assuming that the rent satisfies the condition from the desired value to the value 10,000 yen cheaper, and the other only satisfies the desired value.) The property is searched against the property information table 52, and the result is stored as a search result. Stored in the unit 53.

  The selection unit 54 receives the identification code (ID) of the guest G from the face information etc. processing unit 60 in FIG. 2, and the property information already provided to the guest G of this ID is traced back to the past from the present in the past. It is determined whether or not it is included in the search result storage unit 53, and if it is included, the property information is selected by deleting it from the search result storage unit 53, and the number of properties after selection is displayed via the display control unit 56. Is displayed on the display panel 29.

  If the number of properties in the search result storage unit 53 is less than 5, the process determination unit 57 transfers control to the condition relaxation unit 58, and otherwise passes the search result to the information processing unit 70 in FIG.

  FIG. 8 is a schematic flowchart showing processing by the condition relaxation unit 58.

  (S10) 0 is substituted into a variable i indicating a question type for condition relaxation.

  (S11) If the value of i is 0, 1 or 2, the process proceeds to steps S12, S19 and S21, respectively.

  Here, it is assumed that the answer keywords included in the keyword storage unit 46 are X minutes on foot, the nearest Y station, and rent Z 10,000 yen.

  (S12) 1 is added to X.

  (S13) i = 0, relaxation value V = X, Y is called as an argument, and the question unit 59 is called. In response to this, if i = 0, the question unit 59 is “X minutes on foot from the nearest Y station. The character string of the question “What about distance?” (The variables Y and X given the above arguments) is converted into speech data and supplied to the D / A converter 281. As a result, the audio data is converted into analog data, further passes through the low-pass filter 282, high-frequency components associated with analogization are removed, then amplified by the amplifier 283, and the signal is supplied to the speaker 28 and converted into audio, A question is issued.

  (S14) An answer to this question is obtained via the keyword obtaining unit 40. If this answer indicates affirmation, the process proceeds to step S15, and if not, the process proceeds to step S17.

  (S15) The walking time in the keyword storage unit 46 is updated with the relaxation value V and the search start SEARCH is activated. As a result, the search in the information search unit 51 is performed again as described above, and the number N of properties on the display panel 29 is updated.

  (S16) If N ≧ 5, the process in the condition relaxation unit 58 is terminated, and if not, the process returns to step S11.

  (S17) Increment the value of i by 1 and return to step S11.

  (S18) The nearest station Y is changed (shifted) from the station to the nearest station in time or space.

  (S19) The question part 59 is called with i = 1 and the relaxation value V = Y as arguments. The question part 59 responds to this, and if i = 1, the question “How about the nearest station Y?” A character string obtained by giving the argument to the variable Y) is converted into voice data and supplied to the D / A converter 281. As a result, questions after relaxing the conditions of the nearest station are issued. Next, the process proceeds to step S14.

  (S20) 1 is added to Z.

  (S19) The question part 59 is called by using i = 2 and the relaxation value V = Z as an argument. The question part 59 responds to this, and if i = 2, the question is "How about rent Z 10,000 yen?" A character string (which is obtained by giving the argument to the variable Z) is converted into voice data and supplied to the D / A converter 281. As a result, questions after relaxation of rent conditions are issued. Next, the process proceeds to step S14.

  In this way, a question for relaxing the condition is automatically issued, and when the guest G gives an affirmative answer, a re-search is performed under the relaxed condition. It is repeated until it disappears.

  FIG. 9 is a schematic block diagram illustrating a configuration example of the face information etc. processing unit 60.

  In the face information processing unit 60, video data from the video camera 26 is stored in a temporary storage unit (frame memory) 61. The guest search / registration unit 62 extracts the face image of the guest G from the frame image in the temporary storage unit 61 after the activation of the listening start START. The similarity with each face image stored in the folder is calculated. The face images in the folder are part of the personal information database. Further, the guest search / registration unit 62 checks whether or not there is a similarity that is equal to or higher than the set value, and if it exists, the face image having the highest similarity is updated with the currently acquired one and must exist. For example, a value obtained by adding 1 to the maximum ID (serial number) assigned to the guest G is used as the guest G ID, and is stored together with the face image. For example, as shown in FIG. 10D, the guest G ID is added to the file name of the face image and stored in the predetermined folder.

  On the other hand, the feature amount extraction unit 64 determines a feature amount for determining the gender and age group of the guest G from the entire image of the guest G including the face image to be registered, for example, whether or not a tie is used. Whether or not the skin is worn, whether or not the lipstick is put on, and the luminance distribution of the skin are acquired. The classification unit 65 estimates the gender and age group of the guest G based on the feature amount, and the estimated gender is stored in the personal information table 63 that is a part of the personal information database as shown in FIG. And a row including the age group and the ID.

  When the guest G replies that he / she wants to obtain the search result by e-mail in response to the question of the host H, the two-dimensional code (not shown) attached to the image forming apparatus 10 is imaged by the camera of the mobile phone. An e-mail including a guest ID is transmitted to an address included in the two-dimensional code. This guest ID is displayed on the display panel 29. The output medium determining unit 66 acquires the email address of the guest G from this mail, stores it in the personal information table 63 in association with the ID, and attaches the output medium to the email attachment instead of the paper medium. And the identification code MEDIAID is output.

  The output format determination unit 67 refers to the manually set format / automatic keyword determination table 68 as shown in FIG. 10B based on the gender and age group of the guest G estimated by the classification unit 65. The output format identification code FMTID is determined. The output format is, for example, to enlarge a character when it is estimated to be an elderly group, or to add a pattern to a frame when it is estimated to be a woman.

  In the format / automatic keyword determination table 68, a keyword to be automatically added according to gender and age group may be described. If there is a corresponding keyword for the current guest G, an automatic keyword adding unit is provided. 69 acquires this as AUTOKW, and stores it in the keyword memory | storage part 46 of FIG. This keyword is, for example, “auto lock” for young women.

  FIG. 11 is a schematic functional block diagram illustrating a configuration example of the information processing unit 70.

  The sorting unit 71 of the information processing unit 70 receives the property information from the processing determination unit 57 of FIG. 7, sorts based on the priority set in the priority table 72, and picks up the top five items. For example, the property information is sorted in ascending order of age, then the same age is sorted in ascending order of walking time to the nearest station, and for those with the same age, the total of deposit and key money is in ascending order. Sort.

  When the sorted property information exceeds five items, the selection unit 73 selects the top five items. If a comment ID is added to the property information for the selected property information, the comment adding unit 74 refers to the comment table 75 in which the comment ID and the comment text are associated with each other, and applies the comment ID. By replacing with, the recommended comment is added to the property information. Recommended comments are, for example, “Relatively shallow. It looks beautiful.”, “Close to the station and convenient for both commuting to school.”, “No key money, 0 deposit, initial moving costs are very small. And so on.

  The output editing unit 76 edits the property information from the comment adding unit 74 based on the format FORMAT determined by the output format determining unit 67 of FIG. 9 and outputs it to the output control unit 80. The output control unit 80 causes the image forming apparatus main body 15 to print the property information or transmits the property information file by electronic mail according to the output medium identification code MEDIAAID determined by the output medium determination unit 66.

  FIG. 12 is a flowchart showing an outline of the operation of the real estate property information providing support device 20 as described above.

  (S30) The host H turns on the power of the image forming apparatus 10.

  (S31) Thereby, monitoring by the video cameras 26 and 27 is started.

  (S32 to S34) When the host H is seated on the chair 12 and the guest G is seated on the chair 13 or 14, the keyword acquisition by the keyword acquisition unit 40 is started.

  (S35, S36) If the answer keyword of the guest G corresponding to the latest question of the host H is included, the corresponding answer keyword in the keyword storage unit 46 is overwritten and updated by this.

  (S37 to S39) Based on the answer keyword stored in the keyword storage unit 46, a real estate property is searched, and those that have been provided in the past are excluded, and the number of cases is displayed on the display panel 29.

  (S40, S41) If the host H keeps the face facing the image forming apparatus 10 for 3 seconds or more, the process proceeds to step S42.

  (S42, S43) If the number N of search results is less than 5, a question for relaxing the condition is issued, and the process returns to step S34.

  (S44) If the number N of cases exceeds 5, the top 5 cases are selected.

  (S46) Sort the property information in order of priority, pick up the top five items, and add recommended comments.

  (S47 to S49) The image forming apparatus 10 prints the property information on paper, and the host H takes the print and presents it to the guest G.

  (S50) If the guest G is not satisfied with the output result, the process returns to step S34.

  (S51 to S53) When the guest G is satisfied with the result, the guest G leaves the chair 13 or the chair 14, and the host H also leaves the chair 12. When a predetermined time, for example, 30 seconds elapses after both of them leave the seat, the listening is finished and the information processing apparatus 21 is turned off.

  In the above, preferred embodiments of the present invention have been described. However, the present invention includes various modifications, and other configurations that realize the functions of the respective components described in the above embodiments are used. However, other configurations that would be conceived by those skilled in the art from these configurations or functions are also included in the present invention.

  For example, the number of directional microphones may be two or more. Further, the speaker identification means is not limited to identifying the voice of the host and the guest based on the output intensity of the plurality of directional microphones. For example, the voice waveform of the host vowel is registered in advance and the current voice Before obtaining the similarity of the waveform to the registered waveform, both waveform intensities are normalized, and the sound of the host and the guest is identified based on the similarity, or the mouth portion of the image changes according to the sound input. May identify the sound of the host and the guest by determining whether it is a host or a guest.

  Further, depending on the installation position of the image forming apparatus 10 with respect to the host H and the guest G, only one video camera may be provided. Further, instead of turning off the information processing apparatus 21 in step S53, the information processing apparatus 21 may be shifted to the sleep state to shorten the startup time.

  The search start determination means may be any means that can determine the search start based on the input of predetermined signal information. For example, the search start determination means can be used for inputting a voice “Let's search” or inputting a predetermined marker image. The configuration may be such that the search start is determined based on the search.

DESCRIPTION OF SYMBOLS 10 Image forming apparatus 11 Table 12-14 Chair 15 Image forming apparatus main body 20 Real estate property information provision assistance apparatus 21 Information processing apparatus 22-25 Directional microphone 221,231,241,251 Amplifier 222,232,242,252 A / D Converter 26, 27 Video camera 28 Speaker 281 D / A converter 282 Low-pass filter 29 Display panel 30 Listening start end determination / output start determination unit 31, 43, 61 Temporary storage unit 32 Chair image storage unit 33 Listening start determination unit 34 Output start determination unit 35 End determination unit 40 Keyword acquisition unit 41 H / G determination unit 42 Addition unit 44 Keyword determination unit 45 Keyword table 46 Keyword storage unit 50 Information search processing unit 51 Information search unit 52 Property information table 53 Search result storage unit 54 Selector 5 Provided data storage unit 56 Number display control unit 57 Processing discrimination unit 58 Condition relaxation unit 59 Question unit 60 Face information processing unit 62 G search / registration unit 63 Personal information table 64 Feature quantity extraction unit 65 Classification unit 66 Output medium determination unit 67 Output format determination unit 68 Format / automatic keyword information table 69 Automatic keyword addition unit 70 Information processing unit 71 Sort unit 72 Priority table 73 Selection unit 74 Comment addition unit 75 Comment table 76 Output editing unit 80 Output control unit START Start listening END End of listening OUT Start of output H Host G Guest

Claims (11)

A keyword storage means for identification in which keywords are registered;
Keyword discriminating means for detecting voice and discriminating whether or not any of the registered keywords is included in the voice;
Search keyword storage means for storing the determined keyword;
A search start determination means;
An information database in which real estate property information is registered;
Search means for searching the information database for information satisfying a condition based on the keyword stored in the search keyword storage means in response to the search start determination means determining the search start;
Output means for outputting information search results in a predetermined format and medium;
In the real estate property information providing support device equipped with
A speaker identification means for identifying the voice of the host and the guest;
In the keyword storage means, a question keyword and an answer keyword for each question type classified by one or more question keywords are registered,
The keyword discriminating unit discriminates which of the registered question keywords is included in the voice of the identified host and the type of the question based on the question keyword. Determining which of the registered answer keywords corresponding to the type of the most recent question is included, and storing the determined answer keyword in the search keyword storage means in association with the type of question;
Real estate property information provision support device characterized by that. The search start determination means includes
A video camera,
A cue recognition unit that processes an image captured by the video camera to determine whether or not the orientation of the face of the host is within a predetermined range, and continuously determines affirmatively for a set time or more to determine a search start ,
The real estate property information provision support apparatus according to claim 1, wherein A personal information database in which the face image and visit date of the guest are registered in association with the guest ID;
If the face image of the guest is extracted from the captured image and the face image is not registered in the personal information database, a guest ID is assigned and the face image and the visit date are associated with the guest ID and the personal A guest registration means for registering in the information database;
Information search result storage means for storing the information search result and its date in association with the guest ID;
If the extracted face image has already been registered in the personal information database, whether or not an information search result for a date within a predetermined period from the current date is stored in the information search result storage means for the guest ID If the determination result is affirmative, a search result selection unit that excludes the stored information search result from the current information search result;
The real estate property information provision support device according to claim 2, further comprising: The personal information database further includes the sex and age of the guest as registration items,
Automatic determination information storage means in which automatic keywords associated with gender and age group are selectively registered in advance;
Guest classification means for estimating the guest's gender and age group based on the guest image feature amount and registering the result in the personal information database;
Based on the estimated gender and age group, automatic determination means for reading out the corresponding automatic keyword from the automatic determination information storage means, and adding the automatic keyword to the search condition in the search means;
The real estate property information provision support device according to claim 2 or 3, further comprising: In the automatic determination information storage means, a format identification code associated with the sex and age group is registered in advance,
The automatic determination means further reads out a corresponding format code from the automatic determination information storage means based on the estimated gender and age group, and provides the same to the output means as an identification code of the predetermined format.
The real estate property information provision support apparatus according to claim 4, wherein An email receiver;
If a guest ID is included in the email from the guest, the email address associated with the email is registered in the corresponding guest ID row in the personal information database, and the search result information is email Output medium determining means for providing the output means with an output medium identification code indicating transmission at
The real estate property information provision support device according to claim 4, further comprising: In the information database, keywords in order of output priority are registered in advance,
Sorting means for sorting the information of the final search result based on the keywords of the output priority order and giving the information to the output means;
The real estate property information provision support apparatus according to claim 4, further comprising: A comment ID is selectively added to the information in the information database,
In the information database, a comment ID and a recommended comment sentence are further registered in association with each other.
If a comment ID is added to the sorted information, recommended comment adding means for replacing the comment text corresponding to the comment ID with the comment ID;
The real estate property information provision support device according to claim 7, further comprising: An image picked up by the video camera to determine whether the host and the guest are present at predetermined positions, respectively.
The real estate property information providing support apparatus according to claim 2, further comprising: The image picked up by the video camera is processed to determine whether the host and the guest are no longer at the predetermined position and whether or not this state has passed for a predetermined period of time. Listening end determination means for ending
The real estate property information provision support apparatus according to claim 9, further comprising: Condition relaxation means for outputting a question in which the condition included in the keyword is relaxed;
When the number of final search results is less than the set value, a process determination unit that operates the condition relaxation unit;
The real estate property information providing support apparatus according to claim 1, further comprising:

Images