JP5466190B2 - Server and recommendation method for recommending application to user - Google Patents

Description

  The present invention relates to a server and method for recommending applications to users.

  With the spread and high performance of communication terminals such as portable information terminals, users can download various applications from a server and cause the communication terminals to perform desired operations. There are numerous applications that can be downloaded. While this can provide a variety of options to the user, it causes a problem that the user is at a loss in determining which to select. For this reason, the server recommends some application to the user and assists the user in selecting an application.

  One way to determine an application to recommend to a user is to use a list of user applications that are similar to the user. For example, user A has downloaded applications 1, 2, 3, and 4 to her communication terminal, user B has downloaded applications 1, 2, P, and Q to her communication terminal, and user C has Suppose that applications V and W have been downloaded to the communication terminal. User B owns the same applications 1 and 2 as user A, but user C does not own any of the same applications as user A. Therefore, users A and B have similar preference trends (preferences), and both are similar. However, users A and C are not similar. As a result of such an analysis, among the applications owned by the user B, the applications P and Q not owned by the user A are recommended to the user A.

  However, if the similarity between users is determined by such simple processing, an application that is not simply owned by the user A is uniformly recommended. For this reason, the problem that the application which is not necessarily useful for the user A will be recommended will arise.

  Patent Document 1 discloses a technique for discriminating users who have similar preferences for music data. Specifically, the invention described in Patent Document 1 is similar to the similarity between users by calculating cosine similarity based on acoustic feature data obtained by vector quantization of user favorite music data for each audio section. Is judged. As described above, it is not easy to find a user who has a similar tendency in application preference using the invention described in Patent Document 1 specialized in acoustic data as it is. Furthermore, when searching for applications, not only user preferences but also the operating environment of the communication terminal must be considered. Therefore, the conventional method has a problem that the server cannot easily and appropriately determine the application recommended to the user.

JP 2007-213401

  An object of the present invention is to enable a server to easily and appropriately determine an application recommended to a user.

A server recommending an application to a user according to an embodiment is:
A receiving unit that receives history information indicating a history of actions performed by the user on the application from each of a plurality of users;
Similar user determination for determining one or more similar users similar to the specific user by comparing the historical information of the specific user with the historical information of other users in response to a request from the specific user And
A recommended application determining unit that determines an application recommended to the specific user from among applications owned by one or more similar users;
A transmission unit that notifies the specific user of the recommended application.

  According to one embodiment, the server can easily and appropriately determine an application recommended to the user.

The figure which shows an example of the system used in an Example. The figure which shows the operation example performed between a communication terminal and an application search server. The figure which shows a mode that the similar users B, C, D, etc. are displayed on the communication terminal of the user A. The figure which shows a mode that the application X, Y, Z etc. of recommendation object are displayed on the communication terminal of the user A. The functional block diagram of a communication terminal. The functional block diagram of an application search server.

  The application search server according to an embodiment determines similarity between users by comparing histories of actions, actions, or operations performed by the users on the application. Since it is based on the history of actions such as installation, uninstallation (deletion), activation, and purchase, it is possible to accurately determine the similarity between users. By selecting an application to be recommended from among applications owned by a user having a high degree of similarity, an application meaningful to the user can be recommended.

  Examples will be described from the following viewpoints.

1． system
2． Example of operation
3． Similarity criteria
3.1 Installation history
3.2 Uninstallation history
3.3 Startup history
3.4 Purchase history
3.5 Combination
Four. Communication terminal
Five. Application search server

<1. System>
FIG. 1 shows an example of a system used in the embodiment. The communication terminals 11 and 12 are connected to the application search server 14 via the network 13. For simplicity of illustration, only two communication terminals are shown, but the number is arbitrary.

  The communication terminals 11 and 12 may be any appropriate device that enables communication with the application search server 14. The communication terminals 11 and 12 may be mobile terminals or fixed terminals, but it is necessary that at least an application can be downloaded from the application search server 14 and executed. The communication terminals 11 and 12 are specifically user devices, mobile phones, information terminals, high-function mobile phones, smartphones, personal digital assistants, stand-alone personal computers, portable personal computers, etc. It is not limited to. The communication terminals 11 and 12 acquire history information, and notify the application search server 14 of the acquired history information regularly or irregularly (as necessary). The history information is stored in the application search server 14. The history information indicates a history of actions or actions performed by the user on the application. The actions performed by the user on the application include, but are not limited to, installing the application, uninstalling the application, starting the application, purchasing the application, and the like.

  The network 13 is a communication network that enables communication between the communication terminals 11 and 12 and the application search server 14. The communication network may be a private network, a public network, or both of them. Alternatively, the communication network may include a secure communication network in which safety for exchanging encrypted signals is ensured.

  In addition to storing and managing applications, the application search server 14 also stores and manages history information of each user of the communication terminals 11 and 12. In addition to the application itself and history information, the application search server 14 manages an application owned by each of a plurality of users for each user. For example, the application search server 14 manages the date and time of installing, uninstalling (deleting), starting and purchasing an application for each application and for each user. Further, the application search server 14 determines an application to be recommended to the user from the stored applications, and recommends it to the users of the communication terminals 11 and 12. Although detailed operations will be described later, in general, the application search server 14 determines a similar user who has a similar preference to the user in response to a request from the user, and recommends it from applications owned by the similar user. Decide which application to use. In FIG. 1, for the sake of simplicity, the application search server 14 is shown as a desktop personal computer, but this is not essential for the present application, and any suitable computer may be used.

<2. Example of operation>
FIG. 2 shows an operation example performed between the communication terminals 11 and 12 and the application search server 14 in the system as shown in FIG. First, the communication terminals 11 and 12 acquire or update their own history information when any action is performed on the application. As described above, the history information indicates a history of actions performed by the user on the application. The actions performed by the user on the application include, but are not limited to, installing the application, uninstalling the application, starting the application, purchasing the application, and the like.

  In step S201, the communication terminals 11 and 12 notify the application search server 14 of the acquired history information. The notified history information is stored in the application search server 14. History information acquired from each of a plurality of users is managed for each user. In the figure, step S201 is shown only once at the beginning of the present operation example, but the notification of the history information to the application search server 14 is appropriately performed separately from the operation described later. That is, the timing at which the communication terminals 11 and 12 notify the history information to the application search server 14 may be performed according to a predetermined cycle, or as necessary (for example, in response to a request from the application search server 14) ) May be done.

  In step S203, a request signal is notified to the application search server 14 from any communication terminal (for example, the communication terminal 11). The request signal is a signal for inquiring of the server what application the server recommends to a specific user (user of the communication terminal 11). Such a request signal is transmitted when, for example, the user of the communication terminal 11 wants to download some application from the application search server 14 or what application is available in the market. This is a case where the user wants to know.

  In step S205, in response to the request signal, the application search server 14 compares the history information of the specific user who is the sender of the request signal with the history information of other users, and thus resembles the specific user. Determine one or more similar users. A similar user is a user whose action history for an application is similar to a specific user. Details of similarity determination criteria for determining similar users will be described later. For convenience of explanation, it is assumed that similar users for a specific user A are users B, C, and D. The degree of similarity between users is, for example, “very high”, “very high”, “high”, “not high”, “low”, “very low”, “very low” It may be expressed as a continuous similarity score, or expressed as a percentage between completely matching (100%) and completely different (0%) May be.

  In step S207, the application search server 14 notifies the specific user A of the determined similar users B, C, and D. On the communication terminal 11 of the specific user A, it is displayed that users having a tendency similar to that of the user A are users B, C, and D.

  FIG. 3 schematically shows that users B, C, and D are displayed as similar users on the screen of the communication terminal 11 of user A. In the case of the illustrated example, the degree of similarity of each user is indicated in stages such as “very high”, “very high”, and “high”. The applications owned by these users are subject to recommendation for user A.

  In step S209 in FIG. 2, the user A selects one of the similar users and notifies the application search server 14 of the selected user. For example, it is assumed that the user A selects and notifies the user B having the highest similarity.

  In step S213, the application search server 14 determines an application recommended to the user A from among the applications owned by the similar user B designated by the user A. For example, it is assumed that applications X, Y, and Z owned by the user B are determined as applications recommended for the user A.

  In step S213, the application search server 14 notifies the user A of the recommended applications X, Y, and Z.

  FIG. 4 schematically shows that the applications X, Y, and Z are displayed as recommendation targets on the screen of the communication terminal 11 of the user A. It should be noted that all applications owned by user B may be recommended to user A, or only applications not owned by user A among applications owned by user B may be recommended to user A.

  In step S215 of FIG. 2, when the user decides to download or purchase a specific application, the communication terminal 11 transmits a download signal designating the application to the application search server 14. In response to this, the designated application is downloaded to the communication terminal 11 in step S217. When the application is downloaded and installed, the installation history information is updated. When an application is uninstalled (deleted), the history information of uninstallation is updated. When the installed application is activated, the activation history information is updated. When a paid application is downloaded, purchase history information is updated.

  In the operation example shown in FIG. 2, a similar user is notified in step S207, and any similar user is designated in step S209. However, the processes in steps S207 and S209 are not essential to this operation example and may be omitted. If omitted, in step S211, the application search server 14 selects a similar user with the highest degree of similarity, and determines a recommended application from among applications owned by the similar user.

<3. Similarity criteria>
Hereinafter, the similarity determination criteria that can be used in step S205 in FIG. 2 will be described. As described above, the application search server 14 compares the history information of the specific user who is the sender of the request signal with the history information of the other users, and thus one similar user similar to the specific user is obtained. Decide above. The history information indicates a history of actions, actions or operations performed on the application by the user. For example, an installation history indicating that the application has been installed, an uninstall history indicating that the application has been uninstalled, and an application has been started. Information such as a startup history indicating that the application has been purchased and a purchase history indicating that the application has been purchased, but is not limited thereto. As described below, the similarity determination between a specific user and another user can be performed based on any history information.

<< 3.1 Installation History >>
When the identifier (ID) of an application installed by a specific user A matches the identifier (ID) of an application installed by another user B, the score of similarity is added, and they do not match It is conceivable not to add points. This is because users who have a common application may be similar. Furthermore, the more applications that are in common, the more likely they are similar. For example, when an application having the same identifier (ID) as an application owned by a specific user A is included, one point may be added to the similarity. For example, if a specific user A has applications 1, 2, 3, and 4 and user B has applications 1, 2, P, and Q, 2 points will be added to the similarity between users A and B. Is done.

  There are a wide variety of applications that can be downloaded from the market, and there are many applications that perform similar functions even if their identifiers (IDs) are different. Therefore, if the similarity is determined only by the identifier (ID), there is a concern that a similar user may be missed. Usually, a category is predetermined for an application. The category indicates a classification of the application, and includes, for example, news, business, education, tools, lifestyle, entertainment, games, travel, and the like, but is not limited thereto. Therefore, if the category of the application installed by a specific user A matches the category of the application installed by another user B, the degree of similarity is added, and if they do not match, it may not be added It is done. This determination criterion may be used instead of the similarity determination based on the identifier (ID), or may be used together with the similarity determination based on the identifier (ID).

  The application is accompanied by an explanatory text or an introduction text. The explanatory text explains the function exhibited by the application. Therefore, it is also conceivable to compare application descriptions. For example, if the entire description of all applications installed by a specific user A matches the entire description of all applications installed by another user B, the degree of similarity is added and they do not match In this case, it is considered that no points are added. In this case, it is desirable to add similarity according to the degree of matching, rather than an alternative process of whether or not to add points. In other words, the degree of similarity depends on whether or not a predetermined ratio of words appearing in the descriptions of all applications owned by a specific user A appears in the descriptions of all applications owned by other users B. It may be determined. The comparison between the explanatory texts is, for example, extracting a predetermined part of speech by morphological analysis of the explanatory texts, deriving a feature vector whose component is determined depending on whether or not it has a predetermined word for each user, This may be done by calculating the cosine similarity of the vector.

By the way, if two users own the same application, one user has only used for a short period of time and the other user has used for a long period of time, these users are not similar It may be. This is because users with different proficiency levels for applications are often not similar. Conversely, if the period of use is comparable, these users are considered similar. In this way, the similarity of the user may be determined according to the length (or proficiency level) of a period in which a similar application is used. For example, when a specific user A has installed an application, it is assumed that another user B has installed an application having the same identifier (ID), category, or description. In this case, if the period from the date when the users A and B are installed to the present time is approximately the same, the degree of similarity is greatly added, and if not, the degree of similarity is not greatly added. For example, it is assumed that a user A installed 6 months ago, a user B installed 4 months ago, and a user C installed 1 month ago. In this case, user A is more similar to user B than user C. Even if the user D has installed two years ago, the similarity with the user A decreases in the order of the user B, the user C, and the user D. That is, the smaller the difference in the period during which the application is used, the higher the degree of similarity score. The score of the similarity between the users A and B from such a viewpoint may be calculated by, for example, 1 / (1+ | D _A −D _B |) or exp (− | D _A −D _B |). Good. In any case, the smaller the difference | D _A −D _B | In this case, D _A represents the period from user A installation to the present, and D _B represents the period from user B installation to the present.

  It is preferable that the length of the usage period of the application is taken into consideration depending on the similarity determination based on the coincidence / mismatch of the identifier (ID), category, or description. For example, the score of the similarity due to the mismatch between the identifier (ID), the category or the description is added to the score of the similarity according to the length of use period when the identifier (ID), the category or the description is matched. Thus, the overall similarity score may be calculated.

<< 3.2 Uninstallation History >>
In the same way as when using the installation history in 3.1, it is also possible to use the uninstallation history. Specifically, if the identifier (ID) of an application uninstalled (deleted) by a specific user A matches the identifier (ID) of an application uninstalled (deleted) by another user B, similar If the degrees are added and they do not match, it is possible that points are not added. This is because users who have deleted common applications may be similar. Furthermore, the more applications that are in common, the more likely they are similar. For example, when an application with the same identifier (ID) as an application deleted by a specific user A is deleted, it is conceivable to add 1 point to the similarity. For example, if a specific user A uninstalls applications 1, 2, 3, and 4 and user B uninstalls applications 1, 2, P, and Q, the similarity between users A and B is 2 points. Is added.

  Similar to the case of using the installation history, even when the uninstallation history is used, the similarity of the user may be determined based on the category or description of the application. Specifically, if the category of the application deleted by the specific user A matches the category of the application deleted by the other user B, the similarity is added, and if it does not match, the score is not added It is possible. This determination criterion may be used instead of the similarity determination based on the identifier (ID), or may be used together with the similarity determination based on the identifier (ID). In addition, if the entire description of all applications deleted by a specific user A matches the entire description of all applications deleted by another user B, the degree of similarity was added and did not match In this case, it is considered that no points are added. In this case, it is desirable to add similarity according to the degree of matching, rather than an alternative process of whether or not to add points.

Furthermore, the similarity of the user may be determined according to the length of the period during which the deleted application has been used. For example, when a specific user A deletes a certain application, it is assumed that another user B deletes an application whose identifier (ID), category, or description matches. In this case, if the period of each user from installation to uninstallation is close, it is possible to add a large degree of similarity, and if it is not close, the degree of similarity may not be greatly added. The score of the similarity between the users A and B from such a viewpoint may be calculated by, for example, 1 / (1+ | D _A −D _B |) or exp (− | D _A −D _B |). Good. In any case, the smaller the difference | D _A −D _B | In this case, D _A represents the period until uninstalled after installation by the user A, D _B represents the period until uninstalled after installation by the user B.

  It is preferable that the length of the usage period of the application is taken into consideration depending on the similarity determination based on the coincidence / mismatch of the identifier (ID), category, or description. For example, the score of the similarity due to the mismatch between the identifier (ID), the category or the description is added to the score of the similarity according to the length of use period when the identifier (ID), the category or the description is matched. Thus, the overall similarity score may be calculated.

<< 3.3 Startup History >>
If the identifier (ID) of an application started by a specific user A matches the identifier (ID) of an application started by another user B, the similarity is added. It is possible not to. This is because users who start common applications may be similar. Furthermore, the more applications that are in common, the more likely they are similar.

  Similar to the case of using the installation history and the uninstallation history, when using the activation history, the similarity of the user may be determined on the basis of the application category and description. This is because there are a wide variety of applications that can be downloaded from the market, and it is feared that a similar user may be missed if the similarity is determined only by the identifier (ID). Specifically, if the category of an application started by a specific user A matches the category of an application started by another user B, a similarity is added, and if they do not match, no points are added It is possible. This determination criterion may be used instead of the similarity determination based on the identifier (ID), or may be used together with the similarity determination based on the identifier (ID). In addition, if the entire description of all applications started by a specific user A matches the entire description of all applications started by another user B, the similarity was added and did not match In this case, it is considered that no points are added. In this case, it is desirable to add similarity according to the degree of matching, rather than an alternative process of whether or not to add points. The comparison between the explanatory texts is, for example, extracting a predetermined part of speech by morphological analysis of the explanatory texts, deriving a feature vector whose component is determined depending on whether or not it has a predetermined word for each user, This may be done by calculating the cosine similarity of the vector.

By the way, when two users are running the same application, if one user is rarely activated and the other user is frequent, these users are not very similar. it is conceivable that. This is because the proficiency level of the application varies greatly. Conversely, if the activation frequencies are similar, these users are considered similar. Thus, the similarity of the user may be determined according to the number of activation frequencies that are the number of activations per unit period. For example, when a specific user A has started an application, it is assumed that another user B has started an application whose identifier (ID), category, or description matches. In this case, if the activation frequencies of the users A and B are close to each other, the degree of similarity is greatly increased, and if not, the degree of similarity is not greatly increased. For example, it is assumed that user A starts 10 times a week, user B starts 4 times a week, and user C starts 100 times a week. In this case, user A is more similar to user B than user C. That is, the similarity is higher as the difference in application activation frequency is smaller. The score of the similarity between the users A and B from such a viewpoint may be calculated by, for example, 1 / (1+ | N _A −N _B |) or exp (− | N _A −N _B |). Good. In any case, the smaller the difference | N _A −N _B | In this case, N _A is the frequency of the user A is activated, N _B represents the frequency at which the user B starts.

  It is preferable to consider the degree of application activation frequency depending on the similarity determination based on the mismatch of the identifier (ID), category, or description. For example, the score of the similarity due to the mismatch of the identifier (ID), category or description is added to the score of the similarity depending on the activation frequency when the identifier (ID), the category or the description is matched. Thus, the overall similarity score may be calculated.

<< 3.4 Purchase History >>
Applications that users can download from the market include both free and paid applications. Usually, users download a paid application more carefully than download a free application. Therefore, when determining the similarity of a user, it may be considered to distinguish between a paid application and a free application. Specifically, if the identifier (ID) of an application purchased by a specific user A matches the identifier (ID) of an application purchased by another user B, the degree of similarity is added and matched. If not, it is possible not to add points.

  Similar to the case of using the installation history, the uninstallation history, and the activation history, also in the case of using the purchase history, the similarity of the user may be determined based on the application category and the explanatory text. Specifically, if the category of the application purchased by a specific user A matches the category of the application purchased by another user B, the similarity is added, and if it does not match, the score is not added It is possible. This determination criterion may be used instead of the similarity determination based on the identifier (ID), or may be used together with the similarity determination based on the identifier (ID). Also, if the entire description of all applications purchased by a specific user A matches the entire description of all applications purchased by another user B, the degree of similarity is judged to be high and they match. If not, it may be determined that the degree of similarity is low. In this case, the degree of similarity varies depending on the degree of matching.

Furthermore, (a) the total number of paid applications, (b) the total amount spent for purchasing applications, (c) the ratio (ratio) of the number of paid applications to the total number of applications, and the like may be used for determining the similarity of the user. For example, when the total number, the total amount, or the ratio is close, it is conceivable that the degree of similarity is greatly increased. The score of the similarity between the users A and B from such a viewpoint may be calculated by, for example, 1 / (1+ | M _A −M _B |) or exp (− | M _A −M _B |). Good. In any case, the smaller the difference | M _A −M _B | In this case, M _A represents the total number of applications purchased by the user A, the total amount spent for purchasing paid applications, or the ratio of the number of paid applications to the total number of owned applications. M _B represents the total number of applications purchased by user B, the total amount spent for purchasing paid applications, or the ratio of the number of paid applications to the total number of owned applications.

  Such total number, total amount, and ratio are preferably considered depending on the similarity determination based on the coincidence / mismatch of the identifier (ID), category, or description. For example, add the score of similarity due to mismatch of identifier (ID), category or description and the score of similarity based on total number, total amount or percentage when identifier (ID), category or description matches By doing so, the score of overall similarity may be calculated.

<< Combination >>
The various similarity determination criteria described in the items of 3.1 to 3.4 may be used alone or in combination of two or more. Specifically, two or more scores of similarities based on the following similarity determination criteria may be combined as appropriate.

A1: Additional points due to mismatched identifiers (IDs) of installed applications
A2: Additional points due to mismatch of installed application category
A3: Additional points based on the degree of agreement between the descriptions of installed applications
A4: Additional points based on whether the usage period is the same in A1 to A3
B1: Additional points due to mismatched identifiers (IDs) of uninstalled applications
B2: Additional points due to unmatched categories of uninstalled applications
B3: Additional points based on the degree of agreement in the description of uninstalled applications
B4: Additional points based on whether the usage period is the same in B1 to B3
C1: Additional points due to mismatched identifiers (IDs) of launched applications
C2: Additional points due to mismatches in the category of the launched application
C3: Additional points based on the degree of coincidence of the description of the launched application
C4: Additional points based on whether the activation frequency is the same in C1 to C3
D1: Additional points due to mismatched identifiers (IDs) of paid applications
D2: Additional points due to the mismatch of paid application categories
D3: Additional points based on the degree of agreement between paid application descriptions
D4: Additional points based on whether the total number of paid applications is the same
D5: Additional points based on whether the total amount spent on paid applications is comparable
D6: Additional points based on whether the ratio of the number of paid applications to the total number of owned applications is comparable <4. Communication terminal>
FIG. 5 shows a functional block diagram of the communication terminal. The illustrated communication terminal is used as the communication terminals 11 and 12 in FIG. FIG. 5 shows processing units or functional elements particularly related to the present embodiment among various functions provided in the communication terminal. As described above, the communication terminal is specifically a user device, a mobile phone, an information terminal, a high-function mobile phone, a smartphone, a personal digital assistant, a stand-alone personal computer, a portable personal computer, or the like. However, it is not limited to these. The illustrated communication terminal includes at least a reception unit 51, a user interface 52, a history information acquisition unit 53, a database 54, a transmission signal generation unit 55, and a transmission unit 56.

  The receiving unit 51 receives a signal from the application search server. Such signals include, for example, applications, recommended application information, user information similar to the user of the communication terminal, and the like.

  The user interface 52 includes a visual and / or audible user interface for the user. Examples of the user interface 52 include a display, a control panel with a keypad, a cathode ray tube (CRT), a liquid crystal display (LCD), an organic EL panel, a touch screen, a keyboard, a mouse, a trackball, a microphone, a speaker, or These are any suitable combinations, but are not limited to these. In one embodiment, the user interface 52 includes a touch screen in which a touch sensitive transparent panel covers the display screen. In this embodiment, the user interface 52 displays a user similar to the user of the communication terminal, an application recommended for the user of the communication terminal, and the like.

  The history information acquisition unit 53 acquires history information. The history information indicates a history of actions, actions, or operations performed by the user on the application. The actions performed by the user on the application include, but are not limited to, installing the application, uninstalling the application, starting the application, purchasing the application, and the like.

  The database 54 stores the history information acquired by the history information acquisition unit 53.

  The transmission signal generation unit 55 creates a signal to be transmitted to the application search server. Such a signal is, for example, a signal for notifying history information, a request signal for requesting information on an application recommended by the server, a signal indicating a similar user designated by a user among one or more similar users, and a recommended signal A signal indicating an application designated by a user among a plurality of applications is created.

  The transmission unit 56 transmits the signal created by the transmission signal generation unit 55 to the application search server.

<5. Application search server>
FIG. 6 shows a functional block diagram of the application search server. The illustrated application search server is used as the application search server 14 of FIG. FIG. 6 shows processing units or functional elements particularly related to the present embodiment among various functions provided in the application search server. The illustrated application search server includes at least a receiving unit 61, a database 62, a similar user determining unit 63, a recommended application determining unit 64, and a transmitting unit 65.

  The receiving unit 61 receives a signal from the communication terminal. Such a signal may include a signal including history information, a request signal for requesting information on an application recommended by the server, a signal indicating a similar user designated by one or more of the similar users, and a plurality of recommended applications. A signal indicating an application designated by the user is included.

  The database 62 stores history information, applications, and information on applications owned by various users. The database 62 may be provided inside the application search server 14, or may be provided outside the application search server 14.

  In response to a request from a specific user, the similar user determination unit 63 compares one specific user similar to the specific user by comparing the specific user's history information with the other user's history information. Decide above.

  The recommended application determining unit 64 determines an application recommended for a specific user from among applications owned by one or more similar users.

  The transmission unit 65 transmits a signal to the communication terminal. Such signals include user information similar to the user who sent the request, recommended application information, an application specified by the user, and the like.

  Although the present invention has been described with reference to particular embodiments, they are merely exemplary and those skilled in the art will appreciate various variations, modifications, alternatives, substitutions, and the like. For example, the present invention may be applied to any appropriate mobile communication system in which a communication terminal can download an application via a network. Although specific numerical examples have been described in order to facilitate understanding of the invention, these numerical values are merely examples and any appropriate values may be used unless otherwise specified. Although specific mathematical formulas have been described to facilitate understanding of the invention, these mathematical formulas are merely examples, unless otherwise specified, and any appropriate mathematical formula may be used. The classification of the examples or items is not essential to the present invention, and the items described in two or more items may be used in combination as necessary, and the items described in one item may be combined with other items. It may be applied (as long as it is not inconsistent) to the matters described. For convenience of explanation, an apparatus according to an embodiment of the present invention has been described using a functional block diagram, but such an apparatus may be realized by hardware, software, or a combination thereof. The software is available on random access memory (RAM), flash memory, read only memory (ROM), EPROM, EEPROM, registers, hard disk (HDD), removable disk, CD-ROM, database, server and any other suitable storage medium May be. The present invention is not limited to the above embodiments, and various modifications, modifications, alternatives, substitutions, and the like are included in the present invention without departing from the spirit of the present invention.

11, 12 Communication terminal
13 network
14 App search server
51 Receiver
52 User interface
53 History information acquisition unit
54 Database
55 Transmission signal generator
56 Transmitter
61 Receiver
62 Database
63 Similar user decision part
64 Recommended app decision section
65 Transmitter

Claims (9)

A receiving unit that receives history information indicating a history of actions performed by the user on the application from each of a plurality of users;
Similar user determination for determining one or more similar users similar to the specific user by comparing the specific user's history information with other user's history information in response to a request from the specific user And
And recommendation application determining section for determining the application to be recommended to the previous Symbol particular user,
Possess a transmission unit for notifying the application that the recommended to the specific user, said transmission unit notifies the one or more similar users to the specific user, the reception unit, wherein one or more Information on similar users specified by the specific user among the similar users is received from the specific user, and the recommended application determination unit is configured to store the applications owned by the similar user specified by the specific user. A server for recommending an application to a user, wherein an application recommended for the specific user is determined . The similar user determination unit has a determination criterion for adding the similarity of the two users when the identifier, category, or description of the application indicated in the history information of the two users match. The server according to claim 1 . The identifier, the closer the length of the category or period in which the description is holding the application match has criteria for adding point the similarity of the two users, according to claim 2 server. 3. The server according to claim 2, wherein the server has a determination criterion for adding the similarity of the two users as the activation frequency of an application having the same identifier, category, or explanatory text is closer. The closer the ratio of the total number of paid applications to the total number of applications purchased by the specific user, the total amount spent purchasing the application by the specific user, or the total number of applications owned by the specific user, the two The server according to claim 2, further comprising a criterion for adding a degree of similarity between the users. Whether or not the explanatory texts match, a predetermined ratio of words appearing in the explanatory texts of all the applications owned by the specific user appears in the explanatory texts of all the applications owned by the other users. The server according to claim 2, which is determined based on whether or not to perform. The recommended application determining unit determines an application that is not owned by the specific user among applications owned by a similar user designated by the specific user as an application recommended to the specific user. The server according to any one of claims 1 to 6 . The history information includes first information indicating the history of installing the application, second information indicating the history of uninstalling the application, third information indicating the history of starting the application, and first history indicating the history of purchasing the application. The server according to any one of claims 1 to 7 , comprising one or more of the four pieces of information. A recommendation method performed by the server to recommend an application to a user ,
Receiving history information indicating a history of actions performed by the user on the application from each of the plurality of users ;
Determining one or more similar users similar to the specific user by comparing the specific user history information with other user history information in response to a request from the specific user ;
Notifying the specific user of one or more similar users;
Receiving, from the specific user, information on a similar user designated by the specific user among the one or more similar users;
Among applications the similar user designated by specific user owns, recommendation and a step of notifying the application that the step recommended to determine the application to be recommended to the particular user to the particular user Method.

Images