1. Field of the Invention
The present invention relates to a data mining system for obtaining data used for market research or the like.
2. Description of the Related Art
When information is to be obtained from a user so as to perform the market research, it is a common practice to ask questions to the user by a direct mail or the like and obtain responses from the user. JP-A-2002-207857 discloses a method in which using a server and a terminal connected over a network, transmission and reception of data such as a text is performed for the user, and questions are asked to the user one by one in response to replies to the questions from the user, thereby gathering information. JP-A-2002-297645 and JP-A-11-296525 disclose a technique by which voice data input to a terminal is recognized for extraction of a broad range of keywords, which will be referred to as KWs below, an appropriate KW is selected from among the extracted KWs, and then information associated with the selected KW is acquired at the terminal.
When the market research is performed by the conventional method, it is difficult to obtain information imparted through word-of-mouth communication, which is information indicating a user's real feeling about a product. It is because if information is acquired by asking the user questions prepared in advance and then getting replies to the questions from the user, the user would become conscious of the existence of the questions or a questioner, so that the user thinks in a way he does not usually take; thus, it becomes difficult to acquire the information under a free-thinking state of the user and the information that reflects user's daily life. The information imparted through word-of-mouth communication can be said to be niche information in view of information distributed through media such as newspapers, magazines, and television. However, since the information imparted through word-of-mouth communication is the information based on personal experiences, it is highly credible and is quickly distributed because it spreads by word of mouth. Thus, it often happens that the information has become an open secret when it gets a lot of attention in the media. Catching such information early and utilizing the information for business can contribute to making a great profit.
Since the other party in a cellular phone call or the like is often a close friend, a sweetheart, or other acquaintance, it may often happen that interesting information or useful information will be conveyed to the other party. Accordingly, if the method is used in which contents of a voice call of the cellular phone or the like are recognized, KWs are extracted from the call, and then the market research using these KWs is performed, information imparted through out-of-mouth communication can also be gathered.
However, use of this method might induce privacy invasion; because it is necessary to perform processing in which voice recognition of entire data in the voice call is performed for conversion into sentences in text format, syntax analysis of the sentences in text format is performed, and then KWs are separated and extracted from the sentences.
An object of the present invention is to gather information imparted through out-of-mouth communication without invading personal privacy. A further object of the present invention is to provide information for early tracking a market trend.
In order to achieve the above-mentioned objects, an information gathering method of the present invention includes:
According to the present invention, information imparted through word-of-mouth communication can be gathered without invading individual privacy. Further, information for early tracking a market trend can be provided.
Other objects, features and advantages of the invention will become apparent from the following description of the embodiments of the invention taken in conjunction with the accompanying drawings.
An embodiment of the present invention will be described below.
A terminal 41 is a cellular phone or the like held by a user. A person who wishes to perform market research to know what design of a household electric appliance should be sold holds an information research server 43. A person who actually performs the market research at the request of the person who holds the information research server 43 holds an information gathering server 42. The relationship between the terminal, the information research sever, and the information gathering server is as shown in
Referring to
The terminal 41 is used by the user, and performs data input and output of voice, a character, an image, and the like. The terminal 41 includes an input unit 2, an output unit 3, a category selection unit 6, a communication unit 10, a KW list holding unit 1, a KW recognition unit 5, a KW counting unit 8, a count data holding unit 9, a clock 4, and a position detection unit 7. The input unit 2 receives an input from the user, and corresponds to a keyboard for inputting a text, a microphone for inputting voice, a telephone mouthpiece, or the like. When the cellular phone is employed as the terminal 41, a unit for receiving a voice call and converting the voice call into a data format that can be processed by a microcomputer or the like corresponds to the input unit 2. The output unit 3 displays and outputs necessary information to the user, and corresponds to a terminal screen or a loudspeaker that outputs voice or the like.
The category selection unit 6 selects which field of information gathering the user participates in, in the information gathering system, according to an input from the input unit 2 and sends category selection information to the information gathering server 42 through the communication unit 10. The category selection unit 6 can also be configured to transmit information useful for category selection such as the information on an age, a gender, and an interested field input in advance to the information gathering server 42 when necessary, in addition to performing processing for sending the category selection information.
According to the AIDCA rule, before buying a product, a customer undergoes following psychological changes: first, he directs attention to the product. Then, he has interest in the product, and creates desire for the product. Then, he has conviction that he can get satisfaction by purchasing the product. Finally, he takes action toward purchasing the product. By narrowing down categories for obtaining information to the categories of user's strong interest as described above, the same effect as the result of research on humans alone, of whom attention and interest are provoked, may be obtained, so that a result close to the result of the market research targeting prospects may be obtained. More specifically, accuracy of information gathering can be more increased than in a case where the information gathering system is used for the market research and the like, and information on the user close to buyers can be obtained.
The communication unit 10 performs communication with a communication unit 24 provided for the information gathering server over the data communication network to transmit or receive data between the terminal and the information gathering server. The KW list holding unit 1 holds a list of KWs for recognition transmitted from the information server 42. The KW recognition unit 5 performs KW recognition of data input to the input unit 2. When the KW recognition unit 5 has detected a KW that matches one of the KWs held in the KW list holding unit 1, the KW recognition unit 5 sends to the KW counting unit 8 information for identifying the detected KW. When the input data is in the format of voice or text, a complete match between the KW in the input data and the KW held in the KW list holding unit 1 is not necessary. Even if there is a difference of one character between the spellings of the KWs or the voice input is not precisely accented, KW detection is performed only if the KWs have been found to be the same through appropriate processing. Addition of such processing places a greater burden on processing for the KW recognition. However, it becomes possible to detect all of information input, being intended by the user as the KW, even if the input KW is misspelled or peculiar. Information to be output from the output unit 3 can also be input to the KW recognition unit 5 as shown in a dotted line from the output unit 3 to the KW recognition unit 5, and KW recognition of the information can also be performed. With this arrangement, KW recognition of the information transmitted to the terminal as well the information sent forth from the user of the terminal can also be performed; information can be thereby gathered from a more extensive information source.
The KW counting unit 8 counts the number of times the KW recognition unit 5 has detected a KW. The KW counting unit 8 counts the number of times each KW held in the KW list holding unit 1 has been detected and increments the value of the detected KW in the count data holding unit 9 for each detection. Every time a specified KW has been detected, the count data holding unit 9 holds data while updating the frequency of occurrence of the KW corresponding to the specified KW. When the period of KW recognition at the terminal 41 is specified, the KW recognition unit 5 performs the KW recognition by referring to the clock 4, thereby performing data gathering. When the specified period has elapsed, the terminal 41 finishes the KW recognition and sends data in the count data holding unit 9 to the information gathering server 42 through the communication unit 10. In an example in
The position detection unit 7 obtains the position of a terminal, and can use a technology such as the GPS (Global Positioning System), for example. Positional information detected by the position detection unit 7 is input to the count data holding unit 9 through the KW counting unit 8. The positional information of the terminal at a point in time when a KW has been detected is held together in the count data holding unit 9. As described above, it is also possible to convert this point information in regard to the terminal into plane information indicating whether the terminal is located in a specified area or not and create an association between the plane information and the detected KW, instead of associating point information. Association using the plane information helps more to reduce the amount of information.
With the above-mentioned arrangement, it becomes possible to associate the position of a terminal at a point in time when a KW has been detected with the KW and hold the positional information. This makes it possible for the information gathering server 42 to create KW occurrence data according to the position or the area where a terminal is located.
The information gathering server 42 transmits KWs to the terminal 41 over the data communication network 207 shown in
As described above, the KW selection unit 21 selects KWs from among KWs held in the KW holding unit 22 according to the category and creates the KW list. Category information is received from the terminal 41 through the communication unit 24. Alternatively, it may also be configured to obtain information on users of the terminals connected to the information gathering server 42 from the user management information holding unit 23, which will be described later, and automatically select an optimal category. The KW list created by the KW selection unit 21 is transmitted to the terminals through the communication unit 24. The data in the count data holding unit 9 obtained by the terminal 41 is input to the DB creation unit 25 through the communication unit 24. Data from a plurality of terminals 41 connected to the information gathering server 42 are thus totalized in the information gathering server 42. Information necessary for the market research such as the occurrence time, occurrence frequency, and occurrence rate of each KW in a specified KW list can be obtained for each KW.
The user management information holding unit 23 stores information on users who use the terminals 41. The information on the users includes ages, genders, hobbies, and interested fields of the users. By using the information on the users, stored in the user management information holding unit 23 and detailing the DB of the KW occurrence frequencies, the DB can accommodate such characteristics as the ages, genders, and interested fields. Thus, it can be expected that the usefulness of the DB be enhanced, as described above.
Further, in order to provide service that grants an incentive for transmission of a result of the KW recognition to the users of the terminals 41, a history of data reception is stored in the user management information holding unit 23 when the data in the count data holding unit 9 has been received through the communication unit 24. By referring to this history, users' interests can be tracked as follows: a user 1 is interested in a baseball team A, a user 2 is interested in jeans, and a user 3 is interested in . . . , and so on.
Further, it is desirable to provide a configuration in which deletion of data that the user has once transmitted is performed if he wishes the deletion so that the deleted data is not included in the data created by the DB creation unit 25.
The information research sever 43 is provided to perform market research on a specific field, and is provided by a market research company, for example. The information research sever 43 includes a research KW creation unit 31, a display unit 32, a communication unit 33, and a research result holding unit 34. The research KW creation unit 31 creates KWs to be stored in the KW holding unit 22 and then sent to the terminal 41 for the KW recognition, and creates an appropriate list of KWs such as the names of products, the names of facilities, and the names of persons to be subject to comparison research, according to the research field. The KW list may be automatically created using a KW automatic gathering program or the like. However, if the KW list is created and then registered by a human such as an operator, a knowhow in the market research can be fully utilized, so that a more-carefully-selected and appropriate KW list can be sometimes created. By manually creating the KW list, keywords can be narrowed down more effectively. Further, information that will become noise for desired data can also be reduced. The result of the market research may include the information that will become noise. However, if the KW list is created manually and the need for processing such as reduction of the information that will become noise is eliminated, it becomes possible to reach a desired result of the market research comparatively early. If a market trend can be identified quickly, it becomes possible for a company that would sell character goods, for example, to determine which character to select in an early stage, which may help in concluding a character use contract or the like.
The research result holding unit 34 receives data on the occurrence times or the occurrence frequencies of the KWs created by the DB creation unit 25 after the information gathering server 42 gathered and totalized data from the terminals 41, and stores the data therein. The data stored in the research result holding unit 34 shows the result of the market research and is used by a method of displaying it on the display unit 32 or the like.
Next, steps of the processing, which are details of the processing, will be described. First, the lists of the KWs to be researched by the market research company or the like are created, and the KW lists are transmitted to the information gathering server from the information research server at step S331. This embodiment assumes that the KW lists are transmitted through communication over the data communication network. However, the KW lists can also be sent using a telephone, by fax, or by mail and the KW lists can be passed to the information gathering server in the form of input by the operator, as described before.
On the other hand, a recognition KW DB is created by the information gathering server using the KW lists at step S321. This embodiment assumes the market research on a plurality of fields. Thus, the recognition KW DB created at step S321 includes a plurality of the lists of the KWs for recognition, which are stored in the KW list holding unit shown in
Next, processing for selecting a KW list for performing the KW recognition is performed. This embodiment shows an example where the user selects the field in which he is interested and the KW recognition for the field is performed.
At step S301, information on KW selection is transmitted from the terminal. As the information on the KW selection, the field of the KW list provided in the information gathering server may be directly specified by the user. Alternatively, source data for selecting the field such as user information and positional information on the terminal may be transmitted. At step S322, the list of the KWs for recognition is created by the information gathering server, using the information on the KW selection transmitted at step S301. This processing is performed by the KW selection unit 21 shown in
The KW list created at step S322 is transmitted to the terminal at step S323. The operation then proceeds to step S302, where the terminal receives the KW list and stores it in the KW list holding unit 1.
Next, the operation at the terminal proceeds to step S303, and confirmation as to the start of research using the KWs is performed. In a scheme where the user acquires specific information from the information freely input by the user, personal information of the user might be gathered, thereby invading the privacy of the user. This system, however, recognizes input information within the range of the KW list alone and researches the occurrence times of the KWs. Thus, there is no danger that other information may be randomly acquired. However, the user may sometimes wish no acquirement of data from transmitted personal information. The operation at step S303 is performed in view of this user's wish, and is performed to confirm to the user whether the research using the KW recognition may be conducted or not before the start of the research.
Referring to
In order to avoid unexpected gathering of data from the information transmitted from the user, it is desirable to show the user that the KW recognition is being performed by display or the like. As shown in
When the period specified for the research has been finished, the operation proceeds to step S307 in
Referring to
When a “yes” button 803 in
On the other hand, the DB creation at step S326 is executed by the DB creation unit 25 in
As subsequent processing, at step S327 for transmitting accompanying information, a content obtained by the user from the research is transmitted to the terminal. The content obtained by the user includes accumulation of points due to the incentive or a communication charge discount, for example, both resulting from transmission of the data on the result of the KW recognition. Alternatively, transmission of information on a KW having a high occurrence frequency or information associated with the KW to the terminal in the research in which the user has participated may be performed.
The KW having the high occurrence frequency is the KW showing a much-talked-about or attractive target at the time of the research. Thus, by transmitting information on the target to the terminal, the user can obtain very fresh and useful information. Specifically, as the KW showing the much-talked-about or attractive target, a restaurant that is popular among office ladies can be pointed out. In this case, users targeted for the research are women in their twenties, and they are given the names of food and restaurants. On the other hand, if the communication content of a specific user frequently includes theme parks as dating spots, introduction of a theme park recommended for dating can also be performed.
Though a plurality of confirmation procedures are provided so as not to invade the privacy of the user as described above, the user, who is accustomed to this information gathering system and has faith in this system may consider these confirmation procedures bothersome. Thus, there is a need for providing a scheme that can set part or all of these confirmation procedures to be omitted.
The terminal receives the accompanying information at step S309 and displays them at step S310. When the research period has been completed, processing at step S328 is executed, so that the result of the research resulting from the KW recognition, held in the DB creation unit 25 in
The information gathering period 401 shows the period for gathering data using the KW recognition, and is set based on the research period specified by the market research company. The field 402 stores information on the field to which the KWs for recognition to be transmitted belong. The terminal extracts information stored in the field 402 as necessary and uses the information for display on the screen when confirmation as to the start of the KW recognition is performed, for example. The incentive 404 is the information indicating the content of the incentive to be paid to the user for information gathering using the KW recognition, and is used for display on the screen when confirmation as to the start of the KW recognition is performed, for example. The encryption code 405 is the encryption key for encrypting data gathered using the KW recognition so that others cannot read and abuse the data when the data is transmitted to the information gathering server. The information source 406 specifies targets for the KW recognition. An example shown in
The positional information addition flag 407 is the flag for storing positional information when a KW has been detected and specifying whether to transmit the positional information to the information gathering server or not. When the positional addition flag 407 is present and the KW is detected, information in the position detection unit 7 in
When transmitted from the information research sever 43 to the information gathering server 42, the data shown in
This scheme will be described with reference to a flowchart shown in
Next, at step S1102, data to be deleted is specified. The screen displayed at the terminal at this point is shown in
When it has been determined that the transmitted data to be deleted is within the gathering period, the data can be deleted by the information gathering server, and the count value for the data can be cancelled from the result of totalization. Thus, the operation proceeds to step S1104, and a message for deleting the data is transmitted to the information gathering server. On the other hand, when the gathering period has elapsed, the data have already been totalized and transmitted to the information research server. Thus, deletion of the data cannot be performed. In this case, a message 1301 to the effect that deletion cannot be performed is displayed, as shown in
When receiving the message transmitted from the terminal at step S1104, processing at step S1111 is performed at the information gathering server to call the data for deletion. During the data gathering period, the transmitted data is stored in the user management information holding unit 23 described with reference to
When the terminal has received the message indicating the deletion completion at step S1105, display of the deletion completion is performed at step S1106. An example of display on the screen at this point is shown in
As described above, the KW recognition is performed using voice or data indicated by characters or the like, commonly input to and output from the terminal, and information is obtained from the KW recognition. Thus, subconscious information of the user can be acquired without making the user conscious of having his information gathered from him. Further, since KWs are selected in advance, and the KW recognition is performed by detecting the KWs matching or associated with the seleted KWs from an input to or an output from the terminal, the burden placed on processing at the terminal is more reduced than in the case where recognition of all KWs is performed. Thus, even the terminal that does not have a high processing capability can perform sufficiently practical recognition in a background such as communication processing. Further, by specifying KWs for recognition in advance, the effect of obtaining privacy information such as user's personal information is achieved. Thus, the effect of increasing the feeling of security of the user whose information is gathered is also achieved.
It should be further understood by those skilled in the art that although the foregoing description has been made on embodiments of the invention, the invention is not limited thereto and various changes and modifications may be made without departing from the spirit of the invention and the scope of the appended claims.
Number | Date | Country | Kind |
---|---|---|---|
2003-177437 | Jun 2003 | JP | national |