KR101428252B1 - Method for task list recommanation associated with user interation and mobile device using the same - Google Patents

Abstract

The present invention relates to a service list recommendation method based on a user interaction action that recognizes repeated interaction patterns of a user and enables a user to predict and recommend an operation list to be performed after a current service list based on the repeated interaction patterns, Lt; / RTI >
The method comprising the steps of: tracking a user's interaction activity to collect at least one service item and then clustering to generate at least one clustering group; And a recommending step of analyzing a user's current interaction operation based on the at least one clustering group and selecting and recommending service items to be used after the service item currently used by the user.

Description

TECHNICAL FIELD [0001] The present invention relates to a service list recommendation method based on user interaction in a mobile device, and a mobile device using the service list recommendation method. [0002]

The present invention relates to a method for analyzing a user's current interaction operation based on an interaction operation pattern of a user and allowing a user to predict and recommend a service item likely to be used later.

People using mobile devices such as smartphones and tablet PCs tend to repeat the same or similar patterns when sequentially selecting and using applications or menus. For example, when a credit card text message is received from a credit card company, a message application activates it, copies the message, activates the household application, and stores the copied message contents as expenditure records Lt; / RTI >

However, existing mobile devices do not recognize the repetitive service usage pattern of the user and can not predict the operation relatedness between the service items such as sending and receiving a text message, dialing, using an application, or using an application menu.

Accordingly, the present invention provides a service list recommendation method based on user interaction in a mobile device that recognizes repetitive interaction patterns of a user and allows a user to predict and recommend service items that are likely to be used in the future based on the repetitive interaction patterns, Equipment.

According to an aspect of the present invention, there is provided a clustering method comprising: clustering for tracking at least one service item by tracking a user's interaction action and then clustering at least one service item to generate at least one clustering group; And a recommendation step of analyzing a user's current interaction operation based on the at least one clustering group and selecting and recommending service items to be used after a service item currently being used by the user, Based service list recommendation method.

Wherein the clustering includes: a service sequence generation step of collecting service items while tracking a user's interaction operation, and generating a service sequence including the acquired service items; And a microcluster generating step of clustering the service sequence to generate a microcluster.

Wherein the service sequence generation step comprises: collecting service items used by the interaction operation when an interaction operation is performed by a user; Confirming that the interaction waiting time has passed the set time when the interaction operation is interrupted; And generating a service sequence including the service items collected so far when the interaction wait time passes the set time.

The service sequence generation step may include generating a service sequence and collecting service items used for an interaction operation when an interaction operation is performed by a user; Adding the collected service item to the service sequence if the collected service item is not included in the service sequence; And if the collected service item is included in the service sequence, delivering the service sequence, then generating a new service sequence and adding the collected service item to the new service sequence.

The microcluster generating step may include: searching for a microcluster to which the service sequence belongs based on a service item included in the service sequence; Adding the service sequence cluster to the searched micro-cluster if the micro-cluster is searched; And if the microcluster is not found, creating a new microcluster and adding the service sequence cluster to the new microcluster.

Wherein the step of searching for a microcluster to which the service sequence belongs comprises comparing a service item included in the service sequence with a service item included in the microcluster and comparing the service item included in the service sequence with a service item included in the service sequence, The microcluster can be searched by the microcluster to which the service sequence belongs.

The searching of the microcluster to which the service sequence belongs may search for a microcluster to which the service sequence belongs based on one of the first service item and the last service item.

In addition, the clustering step may further include a macro cluster generation step of clustering the micro clusters to generate macro clusters.

The macrocluster generation step may include calculating a degree of similarity for each pair of microclusters; A step of sorting the pairs of micro clusters having a degree of similarity higher than a preset similarity threshold value and including overlapping micro clusters in order of degree of similarity; And generating macro clusters by combining the micro cluster pairs according to the sorting result.

The step of calculating the degree of similarity for each of the micro-cluster pairs may include the step of using cosine-like statistics.

Wherein the recommendation step comprises: tracking a user's current interaction action to collect service items; And selecting a micro cluster or a macro cluster including a service item having the highest degree of similarity with the collected service item and recommending a service item having the highest frequency of use among the service items included in the cluster to the user .

The step of recommending the service item to the user may include using a TF-IDF (Inverse Document Frequency) algorithm.

The microcluster may include at least one service item and information on frequency of use per service item.

The macro cluster may include at least one service item and information on a frequency of use per service item.

According to another aspect of the present invention, there is provided an information processing apparatus comprising: a collection unit that collects service items while tracking a user's interaction operation, and generates a service sequence including the acquired service items; A micro-clustering unit for clustering the service sequences to generate micro-clusters; And a recommender for selecting and recommending service items to be used after a service item currently being used by the user by analyzing a user's current interaction operation based on the clustering result.

The mobile device may further include a macro clustering unit for once again clustering the clustering result of the micro clustering unit.

According to the present invention, the current interaction operation of the user is analyzed based on the repeated interaction pattern of the user, and the user can predict and recommend a service item likely to be available later, Making it easier to access service items.

In addition, since the service item recommendation operation can replace the existing menu setting or favorite function, the service item recommendation operation has features that are personalized and actively generated and updated by the user. Therefore, It is possible to ensure improved convenience and usability compared to the favorite function.

1 is a view for explaining a service list recommendation method based on a user interaction operation in a mobile device according to an embodiment of the present invention.
2 is a diagram for explaining a service sequence generation step according to an embodiment of the present invention in more detail.
3 is a diagram for explaining a service sequence generation step according to another embodiment of the present invention in more detail. 4 is a diagram for explaining the microcluster generation step (S130) according to an embodiment of the present invention in more detail.
5 is a view for explaining a service list recommendation method based on a user interaction operation in a mobile device according to another embodiment of the present invention.
6 is a diagram for explaining the macrocluster generation step according to another embodiment of the present invention in more detail.
7 is a diagram illustrating a configuration of a mobile device to which a service list recommendation method based on a user interaction operation according to an embodiment of the present invention is applied.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein.

In order to clearly illustrate the present invention, parts not related to the description are omitted, and like parts are denoted by similar reference numerals throughout the specification.

Furthermore, when a part is referred to as being "comprising" or "comprising" an element, it is understood that it may include other elements as well, without excluding other elements unless specifically stated otherwise.

1 is a view for explaining a service list recommendation method based on a user interaction operation in a mobile device according to an embodiment of the present invention.

The service list recommendation method of the present invention analyzes a user's current interaction operation based on an interaction repetition pattern of a user using a mobile device such as a smart phone, a tablet PC, and the like, To predict and recommend high service items.

For example, the 'service element' in the present invention may be a specific function in each application and an application, and the 'interaction operation' is an action for the user to use various services provided by the mobile device, An action of sequentially using a plurality of functions in a specific application, an action of combining a plurality of applications and a plurality of functions in a specific application, and the like.

As shown in FIG. 1, the service list recommendation method of the present invention includes a clustering step (S100) of collecting at least one service item by tracking a user's interaction operation, clustering at least one service item to generate at least one clustering group And a recommendation step of analyzing a user's current interaction operation based on at least one clustering group when the user currently uses the mobile terminal and selecting and recommending service items to be used after the service item currently used by the user S200).

In the clustering step S100 of the present invention, a service sequence generation step (S110) of collecting service items while tracking a user's interaction operation and generating a service sequence including the obtained service items, a service sequence generation step Clms Page number 11 > (S130) for clustering service sequences generated through the service cluster generation step (S110).

In addition, the user's interaction operation can be performed a plurality of times. In the present invention, a plurality of user interaction operations are classified by various methods, and a service sequence for each of them is generated. For example, a plurality of user interaction actions may be tracked and recorded according to time criteria or service item criteria, and a service sequence corresponding to each of them may be generated.

If a new interaction operation is started by a user in a service sequence generation step (S110), a new interaction history is generated and collection of service items is started, When the waiting time has elapsed, the collection of service items corresponding to the interaction operation is terminated.

On the other hand, in the case of tracking and recording the user interaction operation on the basis of the service item, when a new interaction operation is started by the user in the service sequence generation step (S110), the interaction history corresponding thereto is generated and the service item is collected If the service item is collected again, it is determined that a new interaction operation has started to be performed, and a service item collection operation for a new interaction operation other than the previous interaction operation is performed. In other words, based on the service item reuse record of the developer, different interaction operations are distinguished.

When a service sequence corresponding to each user interaction operation is generated and transmitted through a service sequence generation step (S110), a microcluster generation step (S130) generates a service item included in each service sequence and a service item included in the microcluster And searches the microcluster having the same service item as the service sequence.

At this time, the microcluster may include information on at least one service item and a frequency of use per service item (micro cluster MS = {(e ₁ , n ₁ ), (e ₁ , n ₁ ) e _nms , n _nms )}, e is the service item, and n is the frequency of use of the service item).

The service items (e ₁ to e _nms ) included in the same _microcluster indicate service items that the user _{consecutively} uses to perform the specific operation list, and the frequency of use of service items (n ₁ to n _nms ) Which is used mainly by the user during the similar interaction operation.

If the microcluster having the same service item as the service sequence is found as a result of the search, the microcluster generation step (S130) adds the service sequence to the microcluster, and generates a microcluster according to the service item included in the service sequence So that the frequency of use of each service item is updated. However, if a microcluster having the same service item as the service sequence is not found, a new microcluster is created and a service item included in the service sequence is added to the newly created microcluster.

That is, in the microcluster generation step (S130), service sequences having similar service items are clustered to generate at least one microcluster.

In the recommendation step (S200), based on the clustering result of the clustering step (S100), service items that are likely to be used successively after the current service item are predicted and recommended.

To this end, the current interaction behavior of the user is analyzed and the service items used by the user for interaction with the mobile device are collected. A microcluster including service items having the highest similarity to the service items collected through the TF-IDF (Term Frequency-Inverse Document Frequency) algorithm is selected, and a service having a high use frequency among the service items included in the selected microcluster Recommend item to user.

For reference, TF-IDF is a weighting value used in information retrieval and text mining. It is a statistical numerical value indicating how important a word is in a specific document when there is a document group composed of several documents. It can be used to determine the ranking of search results in a search engine, to obtain similarity between documents, and so on. Accordingly, in the present invention, a document is used as a microcluster and a word is used as a service item, and service items related to a user's current interaction operation or a user's interest are selected and recommended to a user.

FIG. 2 is a diagram for explaining the service sequence generation step (S110) according to an embodiment of the present invention in more detail. The service sequence generation section 120 divides a user interaction history different from each other on a time basis and generates a service sequence thereof Lt; / RTI >

As shown in FIG. 2, when the user starts performing a new interaction operation with the mobile device (S111), it tracks the interaction of the user and starts collecting service items used by the user (S112).

If the user interrupts the interaction using the mobile device for a while and enters the interaction wait state (S113), the interaction wait time is counted to check if the interaction wait time elapses (S114).

If the waiting time of the interaction exceeds the set time, it is determined that the user is no longer using the mobile device, and collection of the service item is terminated (S115). That is, in the present invention, a service item corresponding to an interaction operation performed by a user for a specific time is collected.

Then, a service sequence including the service items collected so far is generated and transmitted to the microcluster generation step S130 (S116).

FIG. 3 is a diagram for explaining the service sequence generation step (S110) according to another embodiment of the present invention in more detail. In this case, when a service sequence is generated for differentiating the user interaction history based on service items .

As shown in FIG. 3, when the user starts performing a new interaction operation with the mobile device (S121), a new service sequence is generated and the user starts to collect service items used during the corresponding interaction operation (S122). Then, it is confirmed whether the collected service item is included in the created service sequence (S123).

If the collected service item is not included in the created service sequence, it is determined that the same user interaction operation (or the same service loop) is still being performed, and the collected service item is added to the created service sequence S124) and continues collection of the service list while confirming whether the collected service item is included in the generated service sequence (S125).

On the other hand, if the collected service item is included in the generated service sequence, it is determined that one user interaction operation is terminated and a new user interaction operation is started, and the generated service sequence is stored in a microcluster creation step (S130) (S126).

Then, a new service sequence is generated, and then the process goes to step S123 to add the collected service items to the new service sequence (S127).

In this way, in the method of FIG. 3, when the previously used service item is used again, it is determined that the previously performed user interaction operation is completed, and a service sequence corresponding to the new user interaction operation is generated. For example, if service items a, b, c, a, b, d, b, and d are sequentially acquired, a service sequence having (a, b, c) two service sequences having a, b, d, b, d can be generated.

As described above, the service sequence of the present invention is a sequence of service items corresponding to one user interaction as described above. When a service item that has already been used by a user is reused, one interaction operation is terminated and a new interaction operation It has features that start.

In the present invention, a service item to be reused by a user is used as a separation point to divide a plurality of interaction operations. In this case, assuming that the user's interaction motion is represented by a graph, the splitting point may be an intersection of different interaction actions or be used as a starting point of a new interaction action.

As a result, the service item to be reused by the user is the center of movement of the user, and there is a high possibility that the service item is frequently used by the user, so that the service item to be reused by the user is provided as the first service item of the service sequence .

In addition, if necessary, the service item to be reused by the user may be used as a separation point, and the service item may be provided as the last service item of the service sequence.

4 is a diagram for explaining the microcluster generation step (S130) according to an embodiment of the present invention in more detail.

First, a service item included in the service sequence provided in the service sequence generation step (S110) and a service item included in the microcluster are compared with each other to search for a microcluster to which the service sequence belongs (S131).

In step S131, when the service sequence is generated according to the time reference, the service item included in the service sequence is compared with all of the service items included in the microcluster, and the service item of the service sequence is compared with a predetermined ratio (for example, 50%) or more. On the other hand, when the service sequence is generated based on the service item reference, only the first service item (or the last service item) is compared to determine the microcluster to which the service sequence belongs. This is because, in the case of a service item criterion, the first service item (or last service item) is used as a separation point for a different user interaction history.

If a microcluster to which the service sequence provided in the service sequence generation step (S110) is found (S132) as a search result in step S131, the service sequence is added to the microcluster (S133). For example, if a service sequence including service items (a, b, c) is composed of {(a, 3), (b, 2), (c, 1) The microcluster is updated to {(a, 4), (b, 3), (c, 2), (d, 2) d, 1)}, the microcluster may be updated to {(a, 4), (b, 1), (c, 2), (d, 1)}.

On the other hand, if the corresponding microcluster is not found (S123) as a result of the search in the step 131, a new microcluster is created (S124) and the service sequence provided in the service sequence generation step (S110) is added to the new microcluster (S125). (A, 1), (b, 1), (c, 1)} are generated in a microcluster corresponding to a service sequence including service items (a, b, c) (B, 1), (d, 1), and the microcluster is generated in the microcluster corresponding to the service sequence including the service items of (b, d).

In the above embodiment, the service sequence is clustered to generate a microcluster, and then a recommendation list to be provided to the user is generated based on the generated service cluster. However, if the number of microclusters becomes too large, There may be a problem.

In the present invention, when the number of micro clusters becomes too large, the micro clusters are clustered once again as shown in FIG. 5, and a recommendation list to be provided to the user is generated based on the clusters. So as to prevent the degraded problem in advance.

5 is a view for explaining a service list recommendation method based on a user interaction operation in a mobile device according to another embodiment of the present invention.

 As shown in FIG. 5, the clustering step S100 according to another embodiment of the present invention again tracks a user's interaction operation to collect a plurality of service items, and generates a service sequence including the collected service items A microcluster generation step S130 for generating a microcluster by clustering the service sequences generated through the service sequence generation step S110 and the service sequence generation step S110, similarity metric) to generate a macro cluster again (S150).

In the macrocluster generation step (S150), the similarity between the two microclusters is expressed as "A · B / || A || B ||, where A is a vector of the first microcluster, and B is a vector of the second microcluster Quot ;, and determines that the first and second microclusters are similar to each other only when the calculated cosine similarity is greater than a predetermined similarity threshold.

Hereinafter, the macrocluster generation step (S150) will be described in more detail with reference to FIG.

First, one microcluster pair (i.e., two microclusters) of a plurality of microclusters is selected (S151), and the degree of similarity of the microcluster pairs is calculated using the cosine similarity statistics (S152).

If there are other microcluster pairs other than the selected microcluster pair in step S151 (S153), a new microcluster pair is selected and then the process goes back to step S152 to calculate the similarity of the pair of microclusters.

On the other hand, if the degree of similarity to all of the microcluster pairs has been calculated, the microcluster pairs having the degree of similarity higher than the preset similarity threshold and including the overlapping microclusters are sorted in order of similarity (S154) And the like are combined with each other to generate macro clusters (S155).

(B, 2), (b, 2), (c, 1), (d, 1) (c, 1)}, F: {(f, 2), (c, 1)} and C: {(c, 1) (a, 1), (b, 1) (x, 1)} are present in the present invention, in the present invention, the distance between A and C, between B and A, between E and F, between F and E, A is calculated. (A, C) = 0.79, (C, A) = 0.79 (A, C) = 0.79, and the microcluster pairs having overlapping microclusters are arranged in order of degree of similarity. , (B, A) = 0.67} and {(E, F) = 0.71, (F, E) = 0.71}. (D, 3), (c, 2), (x, 1)}, {E, F} = {(e, 2), (f, 4), (c, 2), (g, 1)}.

In the recommendation step S200, the service list corresponding to the user's current interaction operation can be obtained using only one macro cluster, rather than a plurality of microclusters, so that the operation efficiency of the recommendation step S200 is improved.

In the above embodiment, the microcluster step is performed basically and the macrocluster step is selectively performed. However, if necessary, the macrocluster step may be basically performed to selectively perform the macrocluster step.

7 is a diagram illustrating a configuration of a mobile device to which a service list recommendation method based on a user interaction operation according to an embodiment of the present invention is applied.

7, the mobile device of the present invention analyzes the current interaction operation based on the interaction pattern of the user in addition to the existing main control unit 100 and the user interface unit 200, A service list recommendation control unit 300 for selecting service items to be used after the items, and creating and recommending a service list including the selected service items.

The service list recommendation control unit 300 includes a collection unit 310 for collecting service items while tracking a user's interaction operation and generating a service sequence including the obtained service items, A macro clustering unit 330 for clustering service sequences on the basis of similarity to generate micro clusters, a macro clustering unit 330 for once again clustering micro clusters using cosine similarity metrics to generate macro clusters, A DB (Data Base) 340 for storing clustering results of the micro clustering unit 320 and the macro clustering unit 330, a clustering result stored in the DB 340 (if a macro cluster is stored, Based on the user's current interaction Less analysis by the user is configured to include a recommendation unit 350, for predicting and the like with the high potential service items available for subsequent service items currently used (that is, the list of services). However, the macro clustering unit 330 need not necessarily be provided, and the number of micro clusters may be excessively increased or may be optionally provided according to the needs of the user.

Then, the main control unit 100 displays the service list corresponding to the user's current interaction operation on the current service providing screen through the user interface 200, thereby allowing the user to move out of the current service providing screen, It is possible to check and select the service items to be used on the current service providing screen without having to newly enter the function selecting screen.

For example, assuming that user A has an interaction history of repeatedly connecting Facebook, Cyworld and Twitter, the mobile terminal of the present invention can display the Facebook service screen Cyworld, Twitter By opening a pop-up window recommended for connection, user A can instantly move from the Facebook service screen to the cyworld or Twitter service screen.

If the user A receives a message of the user B and assumes that the user A has an interaction history to promptly reply to the message of the user B, the mobile terminal of the present invention displays a reply message creation menu

By recommending the creation of the reply message immediately after receiving the message of the user B, the user A can more easily perform the reply to the message of the user B.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Modification is possible. Accordingly, the spirit of the present invention should be understood only in accordance with the following claims, and all of the equivalent or equivalent variations will fall within the scope of the present invention.

Also, the service list recommendation method based on the user interaction operation according to the present invention can be implemented as a computer-readable code on a computer-readable recording medium. A computer-readable recording medium includes all kinds of recording apparatuses in which data that can be read by a computer system is stored. Examples of the recording medium include a ROM, a RAM, a CD ROM, a magnetic tape, a floppy disk, an optical data storage device, a hard disk, a flash drive and the like, and also a carrier wave . The computer-readable recording medium may also be distributed over a networked computer system so that computer readable code can be stored and executed in a distributed manner.

Claims (17)

CLAIMS What is claimed is: 1. A clustering method comprising: a clustering step of tracking at least one service item by clustering at least one service item by tracking a user's interaction; And
And a recommending step of analyzing a user's current interaction operation based on the at least one clustering group and selecting and recommending service items to be used after a service item currently used by the user,
The clustering step
A service sequence generation step of collecting service items while tracking a user's interaction operation and generating a service sequence including the collected service items; And
Clustering a service sequence to create a microcluster, wherein the service cluster recommendation method is based on a user interaction operation in a mobile device. delete The method of claim 1, wherein the service sequence generation step
Collecting service items used by the interaction operation when an interaction operation is performed by a user;
Confirming that the interaction waiting time has passed the set time when the interaction operation is interrupted; And
And generating a service sequence including the service items collected so far when the interaction wait time elapses the set time, based on the user interaction action in the mobile device. The method of claim 1, wherein the service sequence generation step
When an interaction operation is performed by a user, generating a service sequence and collecting service items used for an interaction operation;
Adding the collected service item to the service sequence if the collected service item is not included in the service sequence; And
Generating a new service sequence and adding the collected service item to the new service sequence after delivering the service sequence if the collected service item is included in the service sequence; Based service list recommendation method. 2. The method of claim 1,
Searching for a microcluster to which the service sequence belongs based on a service item included in the service sequence;
Adding the service sequence cluster to the searched micro-cluster if the micro-cluster is searched; And
Generating a new microcluster and adding the service sequence clusters to the new microcluster if the microcluster is not searched; and recommending a service list recommendation based on a user interaction operation in the mobile device. 6. The method of claim 5, wherein searching for a microcluster to which the service sequence belongs comprises:
A microcluster that compares a service item included in the service sequence with a service item included in the microcluster and includes a service item included in the service sequence at a predetermined rate is searched by a microcluster to which the service sequence belongs A service list recommendation method based on user interaction behavior in a mobile device. 6. The method of claim 5, wherein searching for a microcluster to which the service sequence belongs comprises:
A service list recommendation method based on a user interaction operation in a mobile device searching for a microcluster to which the service sequence belongs, based on one of a first service item and a last service item. 2. The method of claim 1, wherein the clustering comprises:
And creating a macro cluster by clustering the micro clusters. The method of recommending a service list based on a user interaction operation in a mobile device. 9. The method of claim 8, wherein the macrocluster generation step
Computing the similarity for each of the pairs of micro-clusters;
A step of sorting the pairs of micro clusters having a degree of similarity higher than a preset similarity threshold value and including overlapping micro clusters in order of degree of similarity; And
And combining macrocluster pairs according to the sorting result to generate a macrocluster. The method of recommending a service list based on user interaction in a mobile device. 10. The method of claim 9, wherein calculating the similarity for each of the microcluster pairs
Cosine similarity statistics are used for the service list recommendation method based on the user interaction action in the mobile device. 9. The method of claim 8,
Collecting service items by tracking the user's current interaction behavior; And
Selecting a micro cluster or a macro cluster including a service item having a highest degree of similarity with the collected service item and recommending a service item having the highest frequency of use among the service items included in the cluster, A service list recommendation method based on user interaction behavior in a device. 12. The method of claim 11, wherein recommending the service item to a user comprises:
Wherein a service list recommendation method based on a user interaction operation in a mobile device is performed using a TF-IDF (Term Frequency-Inverse Document Frequency) algorithm. 9. The method of claim 8,
A service list recommendation method based on a user interaction operation in a mobile device including at least one service item and information on frequency of use per service item. 9. The method of claim 8,
A service list recommendation method based on a user interaction operation in a mobile device including at least one service item and information on frequency of use per service item. A computer-readable recording medium having recorded thereon a program for executing a service list recommendation method based on a user interaction operation in the mobile device according to any one of claims 1 to 14. A collection unit for collecting service items while tracking a user's interaction operation and generating a service sequence including the collected service items;
A micro-clustering unit for clustering the service sequences to generate micro-clusters; And
And analyzing a current interaction operation of the user based on the clustering result and selecting and recommending the service items to be used after the service item currently used by the user,
The micro-
Searching for a microcluster to which the service sequence belongs based on a service item included in the service sequence, adding the service sequence cluster to the searched microcluster when the microcluster is searched, and if the microcluster is not searched A new microcluster is created and the service sequence cluster is added to the new microcluster. 17. The method of claim 16,
And a macro clustering unit for once again clustering the clustering result of the micro clustering unit.

Images