KR20180049852A - News recommedation server and method for recommendation news of using the same - Google Patents

Abstract

A news recommendation server and a news recommendation method using the same are disclosed. According to an exemplary embodiment of the present invention, the news recommendation server includes a data collection unit for collecting web browsing history information from a user terminal, a field ratio calculation unit for calculating a ratio of news fields preferred by a corresponding user based on the collected web browsing history information, and a news recommending unit for recommending news to the user terminal by adjusting at least one of an arrangement order of recommended news fields and the amount of recommended news for each field according to the calculated ratio of the news fields. Accordingly, the present invention can actively recommend the news according to a preference change of the user.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a news recommendation server,

Embodiments of the present invention relate to news recommendation techniques.

As smart devices become popular, the focus of Web services is shifting from PCs (personal computers) to smart devices. This is also true of news services. Also, according to the development of news recommendation system, a user is increasingly desirous to automatically receive recommendation of news in a preferred field. Moreover, although the amount of recommended news is desired to be dynamically adjusted depending on the degree of preference, the existing mobile news recommendation system can not dynamically adjust the amount of preference news according to the user's preference.

Korean Patent Laid-Open Publication No. 10-2016-0104067 (2016.09.02)

The embodiment of the present invention is to provide a technique for recommending the order of the news field and the quantity of each news field by adjusting the user's preference.

A news recommendation server according to an embodiment of the present invention includes a data collection unit for collecting web browsing history information from a user terminal; An area ratio calculation unit for calculating a ratio of news fields preferred by the user based on the collected web browsing history information; And a news recommending unit for recommending news to the user terminal based on the ratio of the calculated news fields.

The news recommending unit may adjust at least one of the arrangement order of the news fields to be recommended and the number of news items recommended by each field according to the ratio of the calculated news fields.

The field ratio calculation unit extracts texts from web pages of each news item accessed by the user terminal based on the web browsing history information, and firstly selects analysis target words based on the part of speech in the extracted text , And can determine the fields of each news item based on the first selected analysis subject word.

The sector ratio calculating unit may assign a weight to each of the first to-be-analyzed analysis subject words, to secondarily select a predetermined number of words having a high weight value among the first to-be-analyzed analyzed words, It is possible to determine the field of each connected news according to the inclusion degree of the secondary selected words.

The news recommendation server may further include a news favorability calculating unit for calculating a user's likelihood of each news provided to each user terminal.

The news favorability calculating unit may calculate a user's likelihood value for the news based on the time the user read the news and the sentence length of the news.

Wherein the data collecting unit collects log data to be connected to the news recommendation server from the user terminal, and the news favorability calculating unit is configured to calculate the news favorability of the news recommendation server from the web pages of the web browsing history information, Extracting a web page having a connection point, and calculating a time of reading the news based on the connection time of the extracted web page and the connection time of the web page accessed next to the extracted web page.

Wherein the news favorability calculating unit calculates the speed at which the user reads the news based on the time at which the user read the news and the length of the sentence of the news, The user can set a high value of the user's favorability for the news.

The news recommender searches for a user having a similar tendency to a news subscription to the user and recommends the news according to the ratio of the news field preferred by the user among the news items that the user has a similar tendency to the retrieved news subscription.

Wherein the news recommendation unit extracts, in the descending order of the news appeal of users who have similar news subscription tendencies, the news that the user has a similar news subscription tendency, and extracts the ratio of the news field preferred by the user It is possible to adjust at least one of the arrangement order of the recommended news fields and the number of recommended news items according to the fields.

A news recommendation method according to an embodiment of the present invention is a method performed in a computing device having a memory storing one or more processors and one or more programs executed by the one or more processors, Collecting browsing history information; Calculating a ratio of news fields preferred by the user based on the collected web browsing history information; And recommending news to the user terminal based on the ratio of the calculated news fields.

The step of recommending the news may include adjusting at least one of a sequence order of recommended news fields and a recommended number of news items for each field according to the ratio of the calculated news fields.

Wherein the step of calculating the ratio of the news fields comprises: extracting a text from a web page of each news accessed by the user terminal based on the web browsing history information; Firstly sorting the analysis target word based on the part of speech in the extracted text; Assigning weights to each of the first to-be-analyzed analyzed words; Secondly selecting a predetermined number of words having the highest weight value among the first to-be-analyzed analyzed words; And determining an area of each connected news item in accordance with the inclusion degree of the secondary selected words on the basis of the text.

The news recommendation method may further include calculating a user's likelihood of each news provided to each user terminal.

The step of calculating the good feeling may calculate a user's likelihood value for the news based on the time the user read the news and the sentence length of the news.

The step of calculating the good feeling may include: collecting log data to be connected to the news recommendation server from the user terminal; Extracting a web page having the same connection point as the point of time when the predetermined news is provided from the log data among the web pages of the web browsing history information; And calculating the time of reading the news based on the connection time of the extracted web page and the connection time of the web page accessed next to the extracted web page.

Calculating the rate of reading the news based on a time when the user read the news and a sentence length of the news; And setting a user's likelihood value for the news to a high value according to the calculated degree of the reading speed of the news close to a preset standard speed.

The step of recommending news includes: searching for a user who has a similar news subscription tendency to the user; And recommending the news according to the ratio of the news field preferred by the user among the news items that the user having a similar news subscription propensity is searched for.

Wherein the step of recommending the news comprises the steps of: extracting, in descending order of news appeal from users who have similar news subscription tendencies and have similar news subscription tendencies, And adjusting at least one of a sequence order of news fields to be recommended and a recommended number of news items for each field based on the ratio of the news fields preferred by the user to the extracted news items.

According to the embodiment of the present invention, the ratio of the news field preferred by the user is calculated using the web browsing history per user, and the news is recommended according to the ratio, thereby actively recommending the news according to the change of the user's preference .

1 is a block diagram illustrating a news recommendation system according to an embodiment of the present invention;
2 is a block diagram illustrating a configuration of a news recommendation server according to an embodiment of the present invention;
3 is a diagram illustrating an example of recommending news according to a ratio of a news field preferred by a user in a news recommendation server according to an embodiment of the present invention;
4 is a view illustrating another embodiment of recommending news according to a ratio of a news field preferred by a user in a news recommendation server according to an exemplary embodiment of the present invention
5 is a flowchart illustrating a news recommendation method according to an exemplary embodiment of the present invention.
6 is a flowchart for explaining a news recommendation method according to another embodiment of the present invention
7 is a block diagram illustrating and illustrating a computing environment including a computing device suitable for use in the exemplary embodiments.

Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. The following detailed description is provided to provide a comprehensive understanding of the methods, apparatus, and / or systems described herein. However, this is merely an example and the present invention is not limited thereto.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail. The following terms are defined in consideration of the functions of the present invention, and may be changed according to the intention or custom of the user, the operator, and the like. Therefore, the definition should be based on the contents throughout this specification. The terms used in the detailed description are intended only to describe embodiments of the invention and should in no way be limiting. Unless specifically stated otherwise, the singular form of a term includes plural forms of meaning. In this description, the expressions " comprising " or " comprising " are intended to indicate certain features, numbers, steps, operations, elements, parts or combinations thereof, Should not be construed to preclude the presence or possibility of other features, numbers, steps, operations, elements, portions or combinations thereof.

In the following description, terms such as " transmission ", " transmission ", " transmission ", " reception ", and the like, of a signal or information refer not only to the direct transmission of signals or information from one component to another But also through other components. In particular, "transmitting" or "transmitting" a signal or information to an element is indicative of the final destination of the signal or information and not a direct destination. This is the same for " reception " of a signal or information. Also, in this specification, the fact that two or more pieces of data or information are " related " means that when one piece of data (or information) is acquired, at least a part of the other data (or information) can be obtained based thereon.

Also, the terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms may be used for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

1 is a block diagram illustrating a news recommendation system according to an embodiment of the present invention.

Referring to FIG. 1, the news recommendation system 100 may include a user terminal 102 and a news recommendation server 104. The user terminal 102 is communicably connected to the news recommendation server 104 via the communication network 150. [ In some embodiments, the communications network 150 may include one or more of the Internet, one or more local area networks, wire area networks, cellular networks, mobile networks, other types of networks, As shown in FIG.

The user terminal 102 may include a mobile terminal, a smart device, a desktop, a laptop, a wearable device, and the like as a communication device capable of web browsing. The user terminal 102 is equipped with a web browser.

The user terminal 102 may transmit web browsing history information to the news recommendation server 104. [ The web browsing history information includes, for example, an ID index of a web page accessed by the user terminal 102, a title (title) of the web page, an address of the web page (i.e., a URL of the web page) Whether or not a favorite is set, and when the web page is visited. The user terminal 102 may transmit the web browsing history information to the news recommendation server 104 every time the user terminal 102 accesses the predetermined web page.

The user terminal 102 can access news and provide a web server (not shown) (which may be a server different from the news recommendation server 104) for providing news. In addition, the user terminal 102 can access the news recommendation server 104 and receive the news of the user's favorite field from the news recommendation server 104.

The news recommendation server 104 may receive web browsing history information from each user terminal 102. The news recommendation server 104 may calculate the ratio of news fields preferred by the user based on the web browsing history information received from each user terminal 102. [ That is, the news recommendation server 104 may analyze the web browsing history information for a predetermined period received from each user terminal 102 and calculate the ratio of the news fields preferred by the user among the preset news fields.

When the user terminal 102 accesses the news recommendation server 104, the news recommendation server 104 can recommend (i.e., provide the user terminal 102 with) the news of the user's favorite field. At this time, the news recommendation server 104 may adjust at least one of the arrangement order of the news fields to be recommended and the number of news items recommended for each field according to the ratio of the news field preferred by the user.

The news recommendation server 104 may calculate the likelihood of users for each news provided by the news recommendation server 104. [ The news recommendation server 104 can calculate the likelihood of the news based on the speed at which the user reads the news. The news recommendation server 104 can generate a recommendation news set (Set) by classifying the news in descending order of popularity for each news field. At this time, the recommended news set (Set) can be targeted to the news having the favorable value of the news equal to or more than the predetermined value.

The news recommendation server 104 can recommend news according to the ratio of the news field preferred by the user among the news items of the users who have similar news subscription propensity to the user. The news recommendation server 104 extracts the news items having high news appeal from the news users who have similar news subscription tendencies and the users who have similar news subscription tendencies among the news items, You can arrange the order of the news field and adjust the quantity of the recommended news for each field.

According to the embodiment of the present invention, the ratio of the news field preferred by the user is calculated using the web browsing history per user, and the news is recommended according to the ratio, thereby actively recommending the news according to the change of the user's preference .

2 is a block diagram illustrating a configuration of a news recommendation server according to an embodiment of the present invention.

Referring to FIG. 2, the news recommendation server 104 may include a data collection unit 111, a sector ratio calculation unit 113, a news favorability calculation unit 115, and a news recommendation unit 117.

The data collection unit 111 may collect web browsing history information from each user terminal 102. The data collection unit 111 may store the collected web browsing history information in a database (not shown). Here, the database (not shown) may be a data storage medium provided in the news recommendation server 104 or a data storage medium provided outside the news recommendation server 104.

The data collection unit 111 may collect log data when the news recommendation server 104 is accessed for each user terminal 102. For example, when the user terminal 102 accesses the news recommendation server 104 and receives news, the data collecting unit 111 acquires the log data (the ID of the corresponding news, the title of the corresponding news, URL, time at which the news was provided, etc.). The data collecting unit 111 may store the collected log data in a database (not shown).

Here, the web browsing history information may be a history of viewing the entire web page according to the web browsing of the user, and the log data may be a history of viewing the web page provided from the news recommendation server 104 by the user.

The sector ratio calculating unit 113 may calculate the ratio of the news sectors preferred by each user based on the web browsing history information collected from each user terminal 102. [ In an exemplary embodiment, the news field comprises eight fields, such as sports, entertainment, science, global, politics, economy, life, . ≪ / RTI > Here, the guitar may correspond to a case where the corresponding web page is only an image, or classification is impossible because login is required.

Specifically, the field ratio calculation unit 113 visits each web page accessed by the corresponding user terminal 102 in the web browsing history of the user terminal 102 for a predetermined period of time and displays the body text (i.e., text) Can be extracted. If the web page accessed by the user terminal 102 is a service requiring login or is difficult to extract a body composed only of images, the field ratio calculation unit 113 extracts a title (i.e., title) of the corresponding web page as text .

The field ratio calculation unit 113 can check the part of words included in the text of the extracted web page using a morphological analyzer (Part of Speech Tagger). The field ratio calculation unit 113 calculates a field ratio of the words included in the text of the web page (hereinafter, referred to as " analysis target word ") except the word whose string length is 1 in the verb, adjective, adverb, The frequency of words can be measured. In other words, words of verbs, adjectives, and adverbs are relatively less information than nouns, and words with a string length of 1 are mostly meaningless or meaningless.

The field ratio calculation unit 113 applies a weight to the words to be analyzed so as to calculate a predetermined number (for example, 1,000) of words (hereinafter, may be referred to as a field calculation basic word) in the descending order of the weight value (Or a vector space) can be constructed by extracting a word vector. In the exemplary embodiment, the field ratio calculation unit 113 may apply a term frequency-inverted document frequency (TF-IDF) weight to each analysis target word. Here, the TF-IDF weight can be expressed as a product of a word frequency (Term Frequency) and an inverse document frequency (Inverse Document Frequency). The term frequency can be the number of times a particular word appears within a text divided by the number of words in the text. Inverse Document Frequency is a value obtained by dividing the total number of texts (N) by the number of texts (n) containing specific words (log (N / n)) .

The field ratio calculation unit 113 constructs a matrix for each text and news field using a word vector composed of a predetermined number of words having a high weight value (field calculation basis words) You can classify the field. That is, the field ratio calculation unit 113 can construct a matrix for each text and news field using word vectors as shown in Table 1.

N_money V_interest rate N_family ... N_football News field Text 1 One 0 0 ... 0 Sports Text 2 One One 0 ... 0 Economy Text 3 0 0 One ... One Life ... ... ... ... ... ... ... Text j One 0 0 ... One Sports

The field ratio calculating unit 113 determines whether the field includes a certain number of field-based words (for example, 1 if the word is included or 0 if the word is not included) And apply it to the machine learning technique to decide which field of each news field is. In other words, when a field containing a certain number of basic words is included in the classified news, it is analyzed by a machine learning technique to classify the field into a corresponding field. Then, by using the mattress, Field can be determined. In this case, there are Naive Bayes, K-Nearest Neighbor, Decision Tree, Artificial Neural Network, and Support Vector Machine.

The sector ratio calculating unit 113 may calculate the texts of each web page accessed by the user terminal 102 for a predetermined period of time based on sports, entertainment, science, global, politics, , Economy, life, and others. ≪ Desc / Clms Page number 2 > Sector ratio calculating unit 113 calculates the sum of frequencies of other fields except for other fields (hereinafter, may be referred to as a sum of frequency of all fields) in the texts of the respective web pages accessed by the corresponding user terminal 102 for a predetermined period of time, Can be calculated. The sector ratio calculating unit 113 can calculate the ratio of each field by dividing the frequency of each field (sports, entertainment, science, global, politics, economy, life) by the sum of frequency of all fields. Sector ratio calculating unit 113 may periodically analyze the web browsing history information collected from each user terminal 102 and update the ratio of each user's favorite news field.

When the news recommendation server 104 provides news to the user terminal 102, the news favorability calculating unit 115 may calculate the user's likelihood of each news. The news favorability calculating unit 115 may calculate the user's liking of the news based on the speed at which the user reads the news.

Specifically, the news favorability calculating unit 115 compares the web browsing history information and the log data collected by the data collecting unit 111 and provides predetermined news from the log data among the web page addresses (URLs) of the web browsing history information It is possible to extract a web page having the same connection point as the one point in time. Here, the address of the extracted web page is the same as the address of the web page of the news provided from the news recommendation server 104 to the user terminal 102 (i.e., the URL of the news in the log data).

The reason why the web page URL of the web browsing history information and the web page URL of the log data are matched based on the connection time is that the web browsing history information collected from the user terminal 102 includes As well as the entire web page access record according to the user's web browsing.

The news favorability calculating unit 115 checks the connection time of the web page accessed next to the extracted web page in the web browsing history information and determines whether the user has selected the web page (i.e., the news recommendation server 104) Quot;) ") can be calculated.

More specifically, the news favorability calculating unit 115 calculates the web page access time T (i) of the news and the access time T (i + 1) of the web page accessed after the web page of the news in the web browsing history information of the user ) (I.e., T (i + 1) - T (i)) can be calculated as the time at which the news is read. At this time, the news favorability calculating unit 115 may classify the news as noise when the time of reading the news exceeds a preset time (for example, 10 minutes). That is, when a user views news via the user terminal 102 (e.g., a smart device), the smart device returns to the home screen (main screen) through the home button rather than terminating the web page or the web browser There are many cases. At this time, since the web page is not terminated but is still waiting in the background, it may take a long time to read the news in spite of the fact that the user is not reading the news, If it exceeds the preset time (for example, 10 minutes), the reading of the news can be classified as noise.

The news favorability calculating unit 115 may calculate the speed of reading the news based on the calculated time of reading the news and the sentence length of the news. At this time, if the sentence length of the news is less than a predetermined sentence length (for example, less than 50 characters), the news favorability calculating unit 115 may classify the read news as noise.

The news favorability calculating unit 115 may calculate the user's liking for the news by comparing the speed of reading the news of the user with a predetermined standard speed (for example, 20 to 30 characters / sec). Specifically, the news favorability calculating unit 115 can increase the likelihood value for the news as the user's speed of reading the corresponding news approaches the predetermined standard speed. In an exemplary embodiment, the range of goodwill values for the news may be set to 0-5. The news favorability calculating unit 115 may obtain the degree to which the speed at which the user reads the news is close to a preset standard speed, using a cosine similarity or the like.

The news favorability calculating unit 115 may store the user's liking of each news provided to the user terminal 102 in a database (not shown). That is, as shown in Table 2, the news favorability calculating unit 115 can store the favorability for each news item in a database.

User ID News ID Favorability One One 521 4.7 2 One 620 3.9 3 One 258 3.4 4 2 521 3 ... ... ... ... n n 521 5

The news recommendation unit 117 can recommend the news of the user's favorite field to the user terminal 102 that logs in and connects to the news recommendation server 104 (that is, provides the news to the user terminal 102). The news recommendation unit 117 may adjust at least one of the arrangement order of recommended news fields and the number of news items recommended by each field according to the ratio of the news field preferred by the user.

For example, if the ratio of news items preferred by the user is sports: global: politics: economy: life: science: entertainment = 1: 4: 4: 4: 2: 7: 2, As shown in FIG. 3, can recommend news to users in the order of science, global, politics, economy, light, entertainment, and sports. The news recommendation unit 117 sets the amount of news to be recommended in accordance with the ratio of the news field preferred by the user (i.e., 1: 4: 4: 4: 2: 7: 2) , 4 politics, 4 economies, 2 life, 7 science and 2 entertainment.

Also, when the ratio of the user's favorite news field is changed to Sports: Global: Politics: Economy: Life: Science: Entertainment = 6: 3: 3: 3: 1: 5: 1, As shown in FIG. 4, the news can be recommended to users in the order of sports, science, global, politics, economy, life, entertainment. The news recommendation unit 117 sets the amount of news to be recommended in accordance with the ratio of the news field preferred by the user (i.e., 6: 3: 3: 3: 1: 5: 1) , 3 politics, 3 economies, 1 life, 5 science, 1 entertainment.

The news recommendation unit 117 may analyze the log data of the user terminal 102 to search for other users who viewed the news similar to the news, and then recommend the news among the news users who are searched. Specifically, the news recommendation unit 117 extracts news that has been searched by other users (i.e., a user whose news subscription tendency is similar to that of the user) from the news, It can be recommended according to the ratio. At this time, the news recommendation unit 117 may recommend users who are similar in news subscription tendency among the news in the news sector preferred by the user in descending order of news appeal.

5 is a flowchart illustrating a news recommendation method according to an exemplary embodiment of the present invention. The method shown in Fig. 5 can be performed, for example, by the news recommendation server 104 described above. In the illustrated flow chart, the method is described as being divided into a plurality of steps, but at least some of the steps may be performed in reverse order, combined with other steps, performed together, omitted, divided into detailed steps, One or more steps may be added and performed.

Referring to FIG. 5, the news recommendation server 104 collects web browsing history information from each user terminal 102 (S 101).

Next, the news recommendation server 104 calculates the ratio of the news sectors preferred by each user based on the web browsing history information collected from each user terminal 102 (S 103).

Next, the news recommendation server 104 adjusts at least one of the order of the recommended news field and the recommended number of news items for each field according to the ratio of the news field preferred by each user (S 105).

6 is a flowchart illustrating a news recommendation method according to another embodiment of the present invention. The method shown in Fig. 6 can be performed, for example, by the news recommendation server 104 described above. In the illustrated flow chart, the method is described as being divided into a plurality of steps, but at least some of the steps may be performed in reverse order, combined with other steps, performed together, omitted, divided into detailed steps, One or more steps may be added and performed.

Referring to FIG. 6, the news recommendation server 104 collects web browsing history information from each user terminal 102 and collects log data for each user terminal 102 (S 201).

Next, the news recommendation server 104 calculates the ratio of the news sectors preferred by each user based on the web browsing history information collected from each user terminal 102 (S 203).

Next, the news recommendation server 104 calculates a user's liking (i.e., news favorability) for each news provided to each user terminal 102 (S 205).

Next, when the predetermined user terminal 102 accesses the news recommendation server 104 (S 207), the news recommendation server 104 searches for users having similar news subscription tendencies (S 209). For example, the news recommendation server 104 may search for users who have similar news subscription tendencies by using the Nearest Neighborhood algorithm.

Next, the news recommendation server 104 recommends the news according to the ratio of the news sector preferred by the user among the news users who have similar news subscription tendencies (S 211). For example, the news recommendation server 104 arranges the order of the news field according to the ratio of the news field preferred by the user among the news items of the news users who have a similar tendency to the news subscription, The quantity can be adjusted. In addition, the news recommendation server 104 extracts news items having high news appeal from users who have a similar news subscription tendency and those users who have similar news subscription tendencies among the news items, You can arrange the order of the news fields according to their ratios and adjust the quantity of news items recommended by each field. At this time, it is possible to extract news in the order of news having the highest likelihood of the news in each field, and to recommend the news according to the quantity.

FIG. 7 is a block diagram illustrating and illustrating a computing environment 10 including a computing device suitable for use in the exemplary embodiments. In the illustrated embodiment, each of the components may have different functions and capabilities than those described below, and may include additional components in addition to those described below.

The illustrated computing environment 10 includes a computing device 12. In one embodiment, computing device 12 may be a user terminal (e.g., user terminal 102). In addition, the computing device 12 may be a server that recommends news (e.g., news recommendation server 104).

The computing device 12 includes at least one processor 14, a computer readable storage medium 16, The processor 14 may cause the computing device 12 to operate in accordance with the exemplary embodiment discussed above. For example, processor 14 may execute one or more programs stored on computer readable storage medium 16. The one or more programs may include one or more computer-executable instructions, which when executed by the processor 14 cause the computing device 12 to perform operations in accordance with the illustrative embodiment .

The computer-readable storage medium 16 is configured to store computer-executable instructions or program code, program data, and / or other suitable forms of information. The program 20 stored in the computer-readable storage medium 16 includes a set of instructions executable by the processor 14. In one embodiment, the computer-readable storage medium 16 may be any type of storage medium such as a memory (volatile memory such as random access memory, non-volatile memory, or any suitable combination thereof), one or more magnetic disk storage devices, Memory devices, or any other form of storage medium that can be accessed by the computing device 12 and store the desired information, or any suitable combination thereof.

Communication bus 18 interconnects various other components of computing device 12, including processor 14, computer readable storage medium 16.

The computing device 12 may also include one or more input / output interfaces 22 and one or more network communication interfaces 26 that provide an interface for one or more input / output devices 24. The input / output interface 22 and the network communication interface 26 are connected to the communication bus 18. The input / output device 24 may be connected to other components of the computing device 12 via the input / output interface 22. The exemplary input and output device 24 may be any type of device, such as a pointing device (such as a mouse or trackpad), a keyboard, a touch input device (such as a touch pad or touch screen), a voice or sound input device, An input device, and / or an output device such as a display device, a printer, a speaker, and / or a network card. The exemplary input and output device 24 may be included within the computing device 12 as a component of the computing device 12 and may be coupled to the computing device 12 as a separate device distinct from the computing device 12 It is possible.

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, . Therefore, the scope of the present invention should not be limited to the above-described embodiments, but should be determined by equivalents to the appended claims, as well as the appended claims.

100: News recommendation system
102: User terminal
104: news recommendation server
111: Data collecting unit
113: field ratio calculation unit
115: news favorability calculating section
117: news recommendation department

Claims (17)

A data collection unit for collecting web browsing history information from a user terminal;
An area ratio calculation unit for calculating a ratio of news fields preferred by the user based on the collected web browsing history information; And
And a news recommendation unit for recommending news to the user terminal by adjusting at least one of an arrangement order of recommended news fields and a recommended number of news items for each field according to the ratio of the calculated news fields.
The method according to claim 1,
The sector ratio calculation unit calculates,
Extracting a text from a web page of each news item accessed by the user terminal based on the web browsing history information, firstly selecting an analysis subject word based on the part of speech in the extracted text, And determines a field of each connected news based on a word to be analyzed.
The method of claim 2,
The sector ratio calculation unit calculates,
A weighted value is assigned to each of the firstly selected analysis target words, and a second predetermined number of words having a high weight value is selected from among the first to be analyzed analyzed words, And determines the field of each connected news according to the inclusion degree of words.
The method according to claim 1,
The news recommendation server,
Further comprising a news favorability calculating unit for calculating a user's likelihood of each news provided to each user terminal.
The method of claim 4,
Wherein the news favorability calculating unit comprises:
And calculates a user's likelihood value for the news based on a time when the user read the news and a sentence length of the news.
The method of claim 5,
Wherein the data collection unit collects log data from the user terminal to access the news recommendation server,
Wherein the news favorability calculating unit is configured to extract a web page having the same access point as the time point when the predetermined news is provided from the log data among the web pages of the web browsing history information, And calculates the time at which the news is read based on the connection time of the next accessed web page.
The method of claim 5,
Wherein the news favorability calculating unit comprises:
The speed of reading the news based on the time when the user read the news and the length of the sentence of the news is calculated and the user's likelihood of the news according to the degree that the calculated speed of reading the news approaches the preset standard speed A news recommender server that sets a high value.
The method of claim 4,
Wherein,
Searches for a user having a similar tendency to a news subscription to the user, and recommends the news according to a ratio of a news field preferred by the user among the news that the user has a similar tendency to the retrieved news subscription.
The method of claim 8,
Wherein,
A user having a similar news subscription tendency and having a similar news tendency to the news, and extracting the extracted news in accordance with the ratio of the news field preferred by the user to the extracted news A news recommendation server that adjusts at least one of a sequence order of fields and a quantity of recommended news items per field.
One or more processors, and
A method performed in a computing device having a memory storing one or more programs executed by the one or more processors,
Collecting web browsing history information from a user terminal;
Calculating a ratio of news fields preferred by the user based on the collected web browsing history information; And
And recommending the news to the user terminal by adjusting at least one of an arrangement order of recommended news fields and a recommended number of news items according to the ratio of the calculated news fields.
The method of claim 10,
Wherein the step of calculating the ratio of the news fields comprises:
Extracting a text from a web page of each news accessed by the user terminal based on the web browsing history information;
Firstly sorting the analysis target word based on the part of speech in the extracted text;
Assigning weights to each of the first to-be-analyzed analyzed words;
Secondly selecting a predetermined number of words having the highest weight value among the first to-be-analyzed analyzed words; And
And determining an area of each connected news item according to the inclusion degree of the secondary selected words on the basis of the text.
The method of claim 10,
The news recommendation method includes:
Further comprising calculating a user's likelihood of each news provided to each user terminal.
The method of claim 12,
Wherein the step of calculating the good feeling comprises:
And calculating a user's likelihood value for the news based on the time the user read the news and the sentence length of the news.
14. The method of claim 13,
Wherein the step of calculating the good feeling comprises:
Collecting log data for accessing the news recommendation server from the user terminal;
Extracting a web page having the same connection point as the point of time when the predetermined news is provided from the log data among the web pages of the web browsing history information; And
And calculating the time of reading the news based on the connection time of the extracted web page and the connection time of the web page accessed next to the extracted web page.
14. The method of claim 13,
Wherein the step of calculating the good feeling comprises:
Calculating a speed of reading the news based on a time when the user read the news and a sentence length of the news; And
And setting a user's preference value for the news to a high value according to a degree of the calculated news reading speed approaching a preset standard speed.
The method of claim 12,
The recommending step may include:
Searching for a user having a similar news subscription tendency to the user; And
And recommending the news according to the ratio of the news field preferred by the user among the news items having the similar news subscription tendency.
18. The method of claim 16,
The recommending step may include:
Extracting, in descending order of news appeal from users who have similar news subscription tendencies and have similar news subscription tendencies, And
Adjusting at least one of a sequence order of recommended news fields and a recommended number of news items for each field based on a ratio of the news fields preferred by the user to the extracted news items.

Images