CN108073606B

CN108073606B - News recommendation method and device for news recommendation

Info

Publication number: CN108073606B
Application number: CN201610995502.1A
Authority: CN
Inventors: 涂畅; 张扬; 王砚峰
Original assignee: Beijing Sogou Technology Development Co Ltd
Current assignee: Beijing Sogou Technology Development Co Ltd
Priority date: 2016-11-10
Filing date: 2016-11-10
Publication date: 2021-12-28
Anticipated expiration: 2036-11-10
Also published as: CN108073606A

Abstract

The embodiment of the invention provides a news recommending method and device and a device for recommending news, wherein the method specifically comprises the following steps: acquiring current input content of a current user; when the current input content meets preset news related conditions, acquiring target news corresponding to the main news keywords, or acquiring the main news keywords and the target news corresponding to the related keywords; the main news keywords are matched with the current input content, the main news keywords are obtained according to historical input behavior data of a plurality of users, and the related keywords correspond to the main news keywords; and recommending the target news to the current user. The embodiment of the invention can improve the accuracy of news recommendation and can meet the personalized information requirements of the current user.

Description

News recommendation method and device for news recommendation

Technical Field

The invention relates to the technical field of information processing, in particular to a news recommending method and device and a news recommending device.

Background

With the rapid development of the internet, the amount of network data is continuously increased, which brings convenience to network users to acquire information and also causes the problem of information overload, and how to quickly and effectively search and locate required information in massive data becomes a prominent problem in the current internet development.

In order to solve the above problems, the existing news website can select current popular news on the homepage or the first page of the channel, and place the news in a relatively striking position to help the user find the interested content, so that the user can be effectively helped to quickly and accurately find the needed resources.

The inventor finds that hot news recommended to different users by the existing news website is the same, however, different information requirements exist for different users due to individual reasons such as hobbies and interests, and thus the hot news recommended by the existing news website cannot meet the personalized information requirements of the users, and the accuracy of news recommendation is low. For example, most users are interested in news of "news event a", but it does not mean that all users are interested in news of "news event a", and a small number of users are interested in other news, for example, news of "olympic games", for example, news of a certain star, etc., if the same popular news is recommended to all users, accurate recommendation cannot be achieved, and personalized information requirements of the users cannot be met.

Disclosure of Invention

In view of the above problems, embodiments of the present invention provide a news recommending method, a news recommending apparatus, and an apparatus for news recommending, which overcome the above problems or at least partially solve the above problems.

In order to solve the above problems, the present invention discloses a news recommendation method, comprising:

acquiring current input content of a current user;

when the current input content meets preset news related conditions, acquiring target news corresponding to the main news keywords, or acquiring the main news keywords and the target news corresponding to the related keywords; the main news keywords are matched with the current input content, the main news keywords are obtained according to historical input behavior data of a plurality of users, and the related keywords correspond to the main news keywords;

and recommending the target news to the current user.

Optionally, the current input content meets preset news-related conditions, including:

the current input content is matched with the main news keywords; or

The current input content is matched with the main news keywords, and the current input content accords with a preset news sentence mode; or

The current input content is matched with the main news keywords, and the current input content is matched with the related keywords corresponding to the main news keywords.

Optionally, mining the main news keywords by:

analyzing the vocabulary attribute of each vocabulary in the historical input behavior data of a plurality of users;

and selecting the vocabulary with the vocabulary attribute meeting the preset attribute condition from the historical input behavior data as a main news keyword.

Optionally, the vocabulary attributes include: at least one of word frequency and number of users.

Optionally, the preset attribute condition includes:

the increasing rate of the word frequency in a first preset time period exceeds a first threshold; and/or

The rate of increase of the number of users over a second preset time period exceeds a second threshold.

Optionally, the method further comprises:

and carrying out noise filtration on the main news keywords in a pseudo-correlation feedback mode.

Optionally, the step of performing noise filtering on the main news keywords in a pseudo-correlation feedback manner includes:

searching according to the main news keywords to obtain a corresponding first news search result;

mining related keywords corresponding to the main news keywords from the first news search result;

searching according to the related keywords to obtain a corresponding second news search result;

and judging whether the main news keywords are noise or not according to the occurrence information of the main news keywords in the second news search result, and filtering the main news keywords if the main news keywords are noise.

Optionally, the related keywords corresponding to the main news keywords are mined by the following steps:

mining related keywords corresponding to the main news keywords from historical input behavior data of a plurality of users; and/or

And searching according to the main news keywords to obtain corresponding news search results, and mining related keywords corresponding to the main news keywords from the news search results.

Optionally, the step of obtaining the target news corresponding to the main news keyword includes:

searching target news corresponding to the main news keywords in a news database; or

And acquiring target news corresponding to the main news keywords from the recommended news list corresponding to the main news keywords.

On the other hand, the invention discloses a news recommending device, which comprises:

the input content acquisition module is used for acquiring the current input content of the current user;

the news acquisition module is used for acquiring target news corresponding to the main news keywords or acquiring the main news keywords and the target news corresponding to the related keywords when the current input content meets preset news related conditions; the main news keywords are matched with the current input content, the main news keywords are obtained according to historical input behavior data of a plurality of users, and the related keywords correspond to the main news keywords; and

and the news recommending module is used for recommending the target news to the current user.

the current input content is matched with the main news keywords; or

Optionally, the apparatus further comprises: the first mining module is used for mining main news keywords;

the first excavation module includes:

the analysis submodule is used for analyzing the vocabulary attributes of all vocabularies in the historical input behavior data of a plurality of users; and

and the selection submodule is used for selecting the vocabulary with the vocabulary attribute meeting the preset attribute condition from the historical input behavior data as the main news keyword.

Optionally, the preset attribute condition includes:

Optionally, the apparatus further comprises:

and the noise filtering module is used for carrying out noise filtering on the main news keywords in a pseudo-correlation feedback mode.

Optionally, the noise filtering module comprises:

the first news searching submodule is used for searching according to the main news keywords to obtain a corresponding first news searching result;

a related keyword mining submodule for mining a related keyword corresponding to the main news keyword from the first news search result;

the second news searching submodule is used for searching according to the related keywords to obtain a corresponding second news searching result;

and the judging submodule is used for judging whether the main news keywords are noise or not according to the occurrence information of the main news keywords in the second news search result, and if so, filtering the main news keywords.

Optionally, the apparatus further comprises: the second mining module is used for mining related keywords corresponding to the main news keywords;

the second excavation module includes:

the first mining submodule is used for mining related keywords corresponding to the main news keywords from historical input behavior data of a plurality of users; and/or

And the second mining submodule is used for obtaining corresponding news search results according to the main news keywords and mining related keywords corresponding to the main news keywords from the news search results.

Optionally, the news acquisition module includes:

the third news searching submodule is used for searching target news corresponding to the main news keywords in a news database; or

And the list acquisition submodule is used for acquiring target news corresponding to the main news keywords from the recommended news list corresponding to the main news keywords.

In yet another aspect, an apparatus for news recommendation is disclosed that includes a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured for execution by the one or more processors includes instructions for:

acquiring current input content of a current user;

and recommending the target news to the current user.

The embodiment of the invention has the following advantages:

according to the embodiment of the invention, whether the news requirements related to the current input content exist in the user is accurately identified through the preset news related conditions, news recommendation is carried out only when the current input content meets the preset news related conditions, and the main news keywords based on the target news to be recommended are matched with the current input content, so that the news requirements related to the current input content can be accurately identified, the target news related to the current input content can be provided for the current user, and the accuracy of news recommendation can be improved.

In addition, the target news recommended by the embodiment of the invention is closely related to the current input content of the current user, and the current input content of different users is different, namely, under the condition that the current input content reflects the individuation characteristics of the current user, the target news recommended by the embodiment of the invention can also conform to the individuation information requirements of the current user.

Drawings

FIG. 1 is a schematic diagram of an application environment of a news recommendation method of the present invention;

FIG. 2 is a flowchart illustrating a first step of a news recommendation method according to a first embodiment of the present invention;

FIG. 3 is a flowchart illustrating steps of a second embodiment of a news recommendation method of the present invention;

FIG. 4 is a block diagram of a news recommender in accordance with an embodiment of the present invention;

FIG. 5 is a block diagram of an apparatus 900 for news recommendation of the present invention; and

fig. 6 is a schematic diagram of a server in some embodiments of the invention.

Detailed Description

In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.

The embodiment of the invention provides a news recommendation scheme, which represents hot news concerned by a plurality of users through a main news keyword and obtains the main news keyword according to historical input behavior data of the plurality of users, so that when the current input content of the current user meets preset news related conditions, the current input content is related to the news, namely, the news requirement related to the current input content of the user can be determined, and the target news corresponding to the main news keyword matched with the current input content can be obtained and recommended to the user; according to the embodiment of the invention, whether the news requirements related to the current input content exist in the user is accurately identified through the preset news related conditions, news recommendation is carried out only when the current input content meets the preset news related conditions, and the main news keywords based on the target news to be recommended are matched with the current input content, so that the embodiment of the invention can accurately identify the news requirements related to the current input content, can provide the target news related to the current input content for the current user, and can improve the accuracy of news recommendation.

In one example of the application of the present invention, most users are used to enter news-related content during chatting (via an instant messenger program), such as: "science retired", "do you know news event B? "," do you know the latest message of news event a "," how many gold pieces of olympic games "," do linden have a game recently ", etc., then the embodiment of the present invention may dig out from the historical input behavior data of a plurality of users: main news keywords (such as "science", "chef", "king treasure", "olympic games", "lindane", etc.) representing news of interest to a plurality of users; in this way, when the current input content of the current user meets the preset news related condition (for example, when the current input content matches with the main news keyword), the target news corresponding to the main news keyword can be obtained and recommended to the current user.

The embodiment of the invention can be applied to Application environments such as news websites, news APP (Application), input method APP, search APP, browser APP and the like so as to improve the accuracy of news recommendation; in addition, the embodiment of the invention can be applied to recommendation scenes such as headline news, home pages, pop-up window recommendation, input method candidate recommendation and the like, and it can be understood that the embodiment of the invention does not limit specific application environments and specific recommendation scenes.

The news recommendation method provided by the embodiment of the present invention can be applied to the application environment shown in fig. 1, as shown in fig. 1, the client 100 and the server 200 are located in a wired or wireless network, and the client 100 and the server 200 perform data interaction through the wired or wireless network.

Specifically, the client 100 may obtain current input content of a current user and send the current input content to the server 200; or, the client 100 may obtain the current input content of the current user, and send the analysis result of the current input content to the server 200 after analyzing the current input content; optionally, the analyzing the current input content may include: parsing the current input content to extract a keyword, wherein the non-keyword such as a help word can be filtered out during the extraction process, for example, extracting "science" from "science retired", and for example, extracting "from" do you know news event B? "extract news event B", etc., it is understood that the embodiment of the present invention does not limit the specific process of analyzing the current input content.

After receiving the current input content or the corresponding analysis result, the server 200 determines whether the current input content or the analysis result meets a preset news related condition, and if so, acquires target news corresponding to a main news keyword matched with the current input content, and further outputs the target news to the client 100;

after receiving the target news sent by the server 200, the client 100 may display the target news to the current user for viewing.

Optionally, the client 100 may be run on an intelligent terminal, and the intelligent terminal specifically includes but is not limited to: smart phones, tablet computers, electronic book readers, MP3 (Moving Picture Experts Group Audio Layer III) players, MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop portable computers, car-mounted computers, desktop computers, set-top boxes, smart televisions, wearable devices, and the like.

Method embodiment one

Referring to fig. 2, a flowchart illustrating steps of a first news recommending method according to an embodiment of the present invention is shown, which may specifically include the following steps:

step 201, acquiring current input content of a current user;

step 202, when the current input content meets the preset news related conditions, acquiring target news corresponding to the main news keywords; the main news keywords are matched with the current input content and are obtained according to historical input behavior data of a plurality of users;

step 203, recommending the target news to the current user.

The embodiment of the invention can be applied to a client or a server;

when the method is applied to the client, the current input content is the content input by the user in the application program through the intelligent terminal, and the client can obtain the target news corresponding to the current input content by executing the step 202 and display the target news to the user through the intelligent terminal;

when the method is applied to a server, the current input content may be content sent by a client, and the server may execute step 202 to obtain target news corresponding to the current input content, and push the target news to the client.

The client of the embodiment of the invention can correspond to any APP running on the intelligent terminal, wherein the APP can be an input method APP, a news APP, a browser APP, an instant messaging APP and the like; the server of the input method APP can collect and maintain historical input behavior data of multiple users, and optionally, the server of the input method APP can also provide an acquisition interface for the historical input behavior data of the multiple users, so that other APPs obtain the acquisition interface for the historical input behavior data of the multiple users through the acquisition interface. The embodiment of the invention is mainly explained by taking an input method APP as an example, and other APPs are mutually referred.

It should be noted that, in the embodiment of the present invention, the current user may be identified by a user ID (Identity) or an equipment ID of the intelligent terminal, that is, the embodiment of the present invention does not require the user to log in the corresponding APP by the user ID, and may also identify different users by the equipment ID.

In practical application, when a current user inputs in an application program of the intelligent terminal, the input method APP can capture corresponding current input content in real time. In consideration of the fact that the user habitually communicates news contents through chatting, the application program environment corresponding to the currently input contents may be an instant messenger program. It is understood that the application program environment corresponding to the current input content is not limited by the embodiment of the present invention, for example, for a user who professionally relates to news or a user who likes to make news comments, the current input content related to news can also be input in a document application program environment such as OFFICE.

In the embodiment of the present invention, the current input content may be one or more words or one or more sentences. Optionally, in the context of an instant messaging program, the current input content may be content to be sent in an edit box of the instant messaging program, or the current input content may be content sent and content received in a chat window of the instant messaging program, that is, the current input content may include: at least one of content to be transmitted, content already transmitted, and content already received. The sent content and the received content are relative to the opposite end of the instant messaging, for example, when the current user chats with the user a, the opposite end of the instant messaging is the intelligent terminal of the user a. In summary, in the embodiment of the present invention, the current input content of the current user may include not only the content input by the current user, but also the content input by an opposite-end user communicating with the current user.

In the embodiment of the present invention, the preset news-related condition may be used to constrain the correlation between the current input content and the news. Optionally, since the news main keyword is obtained according to historical input behavior data of a plurality of users, and may represent news of interest of the plurality of users, the current input content meets a preset news-related condition, which may include: the current input content matches the main news keyword, that is, there is a main news keyword that matches the current input content. Alternatively, all the main news keywords may be stored by the main news keyword list, so that the current input content may be matched with each main news keyword in the main news keyword list. Optionally, the matching of the current input content and the main news keyword may specifically include: the vocabulary in the current input content is the same as or similar to or related to the main news keyword, that is, the vocabulary and the main news keyword in the current input content may be similar words or related words. In practical application, a user may have different multiple names for the same news object, for example, "AlphaGo", "a dog", etc. for the name of "a dog", and "baby", "fool root", etc. for the name of "wangbao", the user may all use multiple programs for the same news object as corresponding related words, and it can be understood that the embodiment of the present invention does not limit specific related words of the main news keyword.

In addition, it is understood that the vocabulary in the current input content is the same as or similar to or related to the main news keyword only as an alternative embodiment for matching the current input content with the main news keyword, and actually, those skilled in the art may also adopt other situations where the current input content matches with the main news keyword according to the actual application requirements, for example, the analysis result of the current input content matches with the main news keyword, and the like.

In an optional embodiment of the present invention, in order to accurately constrain the correlation between the current input content and the news, the matching of the current input content with the preset news-related condition may include: the current input content is matched with the main news keywords, and the current input content accords with a preset news sentence mode. Wherein the preset news sentence pattern can be used for restricting the talking sentences corresponding to the news. The preset news sentence pattern can be obtained according to historical input behavior data of a plurality of users or current users. In an application example of the present invention, it is assumed that the mining mode of the embodiment of the present invention for the user to talk about news from the historical input behavior data of a large number of users is as follows: "news you have heard XXX", "do you know XXX", and assume that embodiments of the present invention mined the main news keyword "a doggie" from historical input behavior data of multiple users, so that once the current user or their opposite user enters "do you hear a news of a doggie? "if the current input content corresponds to the preset news sentence pattern," the news listened through XXX "is met, so that the current input content is considered to be in accordance with the preset news sentence pattern. It is understood that a person skilled in the art may adopt a preset news sentence pattern as required according to actual application requirements, and the embodiment of the present invention does not limit the specific preset news sentence pattern.

In the embodiment of the invention, one or more main news keywords can be mined from historical input behavior data of a plurality of users; the multiple users may be all or part of the users in the whole network, and the input method APP may collect historical input behavior data of each user through the client. Moreover, in order to ensure the timeliness of the main news keyword, the historical input behavior data may be data of a preset time period, and a difference between a starting time of the preset time period and a current time may be smaller than a time threshold, for example, the time threshold may be 2 hours, 12 hours, 24 hours, 48 hours, 72 hours, or the like, so that the timeliness of the main news keyword may be ensured. Of course, for convenience of analysis, the starting time (for example, the historical input behavior data before 100 days) exceeding the time threshold is also within the protection scope of the present invention, and the preset time period corresponding to the historical input behavior data is not limited by the embodiment of the present invention.

In an alternative embodiment of the present invention, the main news keywords may be mined by: analyzing the vocabulary attribute of each vocabulary in the historical input behavior data of a plurality of users; and selecting the vocabulary with the vocabulary attribute meeting the preset attribute condition from the historical input behavior data as a main news keyword. Optionally, the vocabulary attributes include: at least one of word frequency and number of users. The word frequency is also the number of occurrences of a certain vocabulary, and optionally, the word frequency may be a total word frequency or a word frequency in a time unit (e.g., one day); the number of users is also the number of users who input a word, and similarly, the number of users may be the total number of users, or the number of users in a time unit (e.g., one day).

In another optional embodiment of the present invention, the preset attribute condition may specifically include:

The situation that the increasing rate of the word frequency in the first preset time period exceeds the first threshold value can correspond to the situation that the word frequency increases suddenly. For example, the input amount of a certain vocabulary is only about 50000 per day in the past 100 days, and the input amount of the vocabulary suddenly increases today to be 100000, which indicates what is related to the word and possibly news related to the vocabulary.

The increasing rate of the word frequency in the first preset time period exceeding the first threshold may correspond to a case where the word frequency increases gradually. For example, a word in a certain vocabulary in the first preset time period may be considered to be related to news, such as increasing the word frequency day by day, 4w in the past, 5w in yesterday, and 6w in today.

The rate of increase of the number of users over the second preset time period exceeding the second threshold may correspond to a situation where the number of users inputting a certain vocabulary is abruptly increased. For example, if the number of users per day of a certain vocabulary was 5000, and the number of users per day now suddenly becomes 10000, the vocabulary may be considered to be related to news.

The rate of increase of the word frequency in the first preset time period may be represented as: the ratio of the word frequency difference between two adjacent days (the next day and the previous day) to the word frequency of the previous day, and the increase rate of the number of users in the second preset time period may be represented as: the ratio of the difference between the number of users on two adjacent days (the next day and the previous day) to the number of users on the previous day can be understood, and those skilled in the art can adopt a reasonable first threshold and a reasonable second threshold according to the actual application requirements.

In addition, it is understood that the preset attribute conditions are only used as an optional embodiment of the present invention, and actually, a person skilled in the art may also use other preset attribute conditions according to actual application requirements, for example, the word frequency or the number of users exceeds a corresponding threshold, and the embodiment of the present invention does not limit specific preset attribute conditions.

In practical applications, the main news keywords obtained by mining may include noise, that is, some vocabularies may be mistakenly recognized as the main news keywords, but actually do not correspond to a piece of news. In an optional embodiment of the present invention, the main news keyword may be further subjected to noise filtering in a pseudo-correlation feedback manner. The pseudo-correlation feedback mode can be used for performing news search by taking the main news keywords as search words and verifying whether the main news keywords are noise or not according to corresponding first news search results.

In another optional embodiment of the present invention, the step of performing noise filtering on the main news keyword in a pseudo-correlation feedback manner may specifically include: searching according to the main news keywords to obtain a corresponding first news search result; mining related keywords corresponding to the main news keywords from the first news search result; searching according to the related keywords to obtain a corresponding second news search result; and judging whether the main news keywords are noise or not according to the occurrence information of the main news keywords in the second news search result, and filtering the main news keywords if the main news keywords are noise. Optionally, N first news search results arranged in front may be selected, news contents of the N first news search results are segmented, word frequencies of the words are counted, M words with the highest word frequency are selected as related keywords, and then the M related keywords are searched as search words to obtain corresponding second news search results, where M, N are all natural numbers. The occurrence information of the news main keyword in the second news search result may include: the occurrence frequency, the reverse file frequency and the like can judge whether the main news keywords are noise or not according to the occurrence information. For example, if the ratio of the number of times of occurrence of the main news keyword in the second news search result to the number of the second news search result is greater than a third threshold, the main news keyword may be regarded as not being noise, and thus may be retained, otherwise, the main news keyword may be noise.

It can be understood that the above-mentioned way of performing noise filtering on the main news keyword by using pseudo-correlation feedback is only an alternative embodiment, and actually, a person skilled in the art may adopt other ways of performing noise filtering, for example, a way of manual review, a way of user voting, or the like.

It should be noted that, a person skilled in the art may determine specific values of the first threshold, the second threshold, and the third threshold according to time application requirements, for example, the specific values of the first threshold, the second threshold, and the third threshold may be determined through an experimental manner, and the specific values of the thresholds such as the first threshold, the second threshold, and the third threshold and the determination manner thereof are not limited in the embodiment of the present invention.

In an embodiment of the present invention, optionally, the step of obtaining the target news corresponding to the main news keyword may specifically include:

The target news corresponding to the main news keywords searched in the news database can be searched online correspondingly, and timeliness of the target news can be guaranteed. Specifically, news search can be performed by using the main news keywords as search terms through a query interface provided by a news database, and news ranked at the top P can be selected as target news.

And acquiring a corresponding offline search mode of target news corresponding to the main news keyword from the recommended news list corresponding to the main news keyword. That is, in an offline manner, news search may be performed in advance through a query interface provided by a news database with the main news keywords as search terms, so as to obtain a corresponding recommended news list. Step 202 may select, as the target news, the news ranked at the top P from the recommended news list corresponding to the main news keyword, and the offline search mode may improve the acquisition efficiency of the target news compared with the online search mode. It is understood that the number P of target news may be determined by those skilled in the art according to the actual application requirements.

In an application example of the invention, assuming that main news keywords such as ' science ratio ' and ' Chuanshi ' are mined based on historical input behavior data of a large number of users, a chat process is carried out on the current user through an instant communication environment, and if ' science ratio # retires #, a plurality of corresponding target news searched based on ' science ratio ' can be recommended to the current user; if "you # know # chuan teacher # news event #? "several target news searched based on" chef chuaner "are recommended to the current user. Optionally, the current user may directly click a link to view the recommended target news, or may directly share the target news with the peer user who is chatting.

The method and the device can be applied to recommendation scenes such as headline news, home pages, popup recommendation and input method candidate recommendation. For example, in an application environment such as a news website and a news APP, the target news may be presented at a headline position, specifically, after a user has just input content XXX, the news APP may present target news related to XXX in a pop-up manner after the user opens the news APP, or present target news related to XXX at a prominent position such as a headline position. For another example, in an application environment of the input method APP, the target news may be presented around a candidate window of the input method APP, for example, if the user has just input "science # retired #", the input method APP may present the corresponding target news at the first time; or the input method APP can display the target news in a pop-up window mode. It is understood that the embodiment of the present invention does not limit the specific manner of recommending the target news to the current user.

In summary, the news recommendation method of the embodiment of the present invention accurately identifies whether the user has a news requirement related to the current input content through the preset news related condition, and performs news recommendation only when the current input content meets the preset news related condition, and the main news keyword based on the target news to be recommended is matched with the current input content.

Method embodiment two

Referring to fig. 3, a flowchart illustrating steps of a second embodiment of the news recommendation method of the present invention is shown, which may specifically include the following steps:

301, acquiring the current input content of the current user;

step 302, when the current input content meets preset news related conditions, acquiring main news keywords and target news corresponding to the related keywords; the main news keywords are matched with the current input content, the main news keywords are obtained according to historical input behavior data of a plurality of users, and the related keywords correspond to the main news keywords;

step 303, recommending the target news to the current user.

Compared with the first method embodiment shown in fig. 2, the present embodiment may obtain the main news keywords and the target news corresponding to the related keywords, so that more accurate news may be recommended to the current user.

In the embodiment of the present invention, the current input content meets a preset news related condition, which may specifically include any one of the following conditions:

the current input content is matched with the main news keywords; or

The conditions that the current input content is matched with the main news keywords and the related keywords further ensure that the current input content of the current user is related to news on the basis of the main news keywords through the related keywords, so that more accurate news can be recommended to the current user. For example, there are many news related to a main news keyword "chuanshi", and once the current input content of the user simultaneously contains related keywords (such as keywords included in news related to "chuanshi"), the main news keyword + the related keywords may be used as search terms to further narrow the scope of recommended news, that is, to lock news related to the news event B and recommend the news to the user, thereby ensuring the recommendation accuracy.

In an application example of the present invention, it is assumed that, based on historical input behavior data of a large number of users, main news keywords such as "science", "academy", and the like are mined, and related keywords corresponding to "science" are mined: the method comprises the steps of 'retirement', 'competition', and the like, and the relevant keywords corresponding to 'Chuanshi' are excavated, so that in the process that a current user chats through an instant communication environment, if 'science ratio # retired', a plurality of corresponding target news searched by 'science ratio + retired' and 'science ratio + competition' can be recommended to the current user; if "do you know # Chuangshi # News #? "several target news searched based on" chuan shi + related keyword 1 "and" chuan shi + related keyword 2 "are recommended to the current user. It is understood that the related keywords corresponding to the target news may be related keywords included in the currently input content, or the related keywords corresponding to the target news may be related keywords not included in the currently input content, so that the user may be provided with precise and rich news.

In an optional embodiment of the present invention, the step of mining the related keywords corresponding to the main news keywords may specifically include:

the method comprises the following steps of 1, mining related keywords corresponding to main news keywords from historical input behavior data of a plurality of users; and/or

And 2, searching according to the main news keywords to obtain corresponding news search results, and mining related keywords corresponding to the main news keywords from the news search results.

The method 1 may analyze historical input behavior data of a plurality of users, for example, may count words appearing in the historical input behavior data of the plurality of users at the same time as the main news keyword, and select a word having a co-occurrence number greater than a fourth threshold as a related keyword according to the co-occurrence number. Optionally, the vocabulary with co-occurrence times larger than a fourth threshold may be further filtered. For example, for a vocabulary with the co-occurrence number greater than the fourth threshold, whether the vocabulary appears in the news search result corresponding to the main news keyword is judged, if yes, the vocabulary is taken as the relevant keyword, and otherwise, the vocabulary is filtered. For another example, the vocabulary with the co-occurrence number greater than the fourth threshold and the news keyword may be used as search words to perform news search, and whether the corresponding news search result is related to the news search result corresponding to the news keyword is determined, if so, the corresponding news search result is used as a related keyword, otherwise, the related keyword is filtered, and the like.

Mode 2 can obtain a corresponding first news search result according to the main news keyword search; and mining related keywords corresponding to the main news keywords from the first news search result. Optionally, the N first news search results arranged in the front may be selected, the news contents of the N first news search results may be segmented, the word frequency of each word may be counted, and M words with the highest word frequency may be selected as the related keyword.

It can be understood that the above mode 1 and mode 2 are only optional mining modes for the related keywords corresponding to the main news keywords, and in fact, a person skilled in the art may adopt any mining mode according to actual application requirements, for example, a manual mining mode, and for example, a target search word including the main news keywords is selected from the popular news search words, and then, words except for the main news keywords in the target search word are used as the related keywords, and the embodiment of the present invention does not limit the specific mining modes for the related keywords corresponding to the main news keywords.

In summary, the news recommendation method of the embodiment of the invention has the following advantages:

firstly, without manual intervention, automatically mining main news keywords representing hot news concerned by a plurality of users according to historical input behavior data of the users, wherein the main news keywords are different from traditional hot news search words, and the main news keywords of the embodiment of the invention are derived from the historical input behavior data because some news are hot, but the users do not always input related words through an input method APP, so that the method for mining hot news based on the input quantity of words can hardly mine such news, for example, in the process of mining hot news based on the input quantity of words, words with the input quantity arranged at the top P position can be mined as keywords of hot news, and words behind the P position cannot be mined, wherein P is an integer greater than or equal to 1, and P can be 10 and the like; the embodiment of the invention utilizes the rule that most users are used to input the content related to news in the chatting process (through an instant messaging program) to mine the main news keywords, so that the main news keywords conforming to the habits of most users can be obtained, and the coverage rate of the main news keywords representing the popular news can be improved;

secondly, under the condition of no need of manual intervention, mining noise data contained in the obtained main news keywords in a pseudo-correlation feedback mode;

and based on the current input content of the current user, automatically recommending personalized news which accords with the current input content to the user.

In addition, whether the news requirements related to the current input content exist in the user is accurately identified through the preset news related conditions, news recommendation is carried out only when the current input content meets the preset news related conditions, and the main news keywords based on the target news to be recommended are matched with the current input content, so that the news requirements related to the current input content can be accurately identified, the target news related to the current input content can be provided for the current user, and the accuracy of news recommendation can be improved.

And moreover, the main news keywords and the target news corresponding to the related keywords are obtained, so that more accurate news can be recommended to the current user.

It should be noted that, for simplicity of description, the method embodiments are described as a series of motion combinations, but those skilled in the art should understand that the present invention is not limited by the described motion sequences, because some steps may be performed in other sequences or simultaneously according to the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no moving act is required as an embodiment of the invention.

Device embodiment

Referring to fig. 4, a block diagram of a structure of an embodiment of a news recommender according to the present invention is shown, which may specifically include: an input content acquisition module 401, a news acquisition module 402, and a news recommendation module 403.

The input content acquiring module 401 is configured to acquire current input content of a current user;

a news acquisition module 402, configured to acquire target news corresponding to a main news keyword when the current input content meets a preset news related condition, or acquire the main news keyword and the target news corresponding to the related keyword; the main news keywords are matched with the current input content, the main news keywords are obtained according to historical input behavior data of a plurality of users, and the related keywords correspond to the main news keywords;

a news recommending module 403, configured to recommend the target news to the current user.

Optionally, the step of matching the current input content with the preset news-related condition may include:

the current input content is matched with the main news keywords; or

Optionally, the apparatus may further include: the first mining module is used for mining main news keywords;

the first excavation module may include:

Optionally, the vocabulary attributes may include: at least one of word frequency and number of users.

Optionally, the preset attribute condition may include:

Optionally, the apparatus may further include:

Optionally, the noise filtering module may include:

Optionally, the apparatus may further include: the second mining module is used for mining related keywords corresponding to the main news keywords;

the second excavation module may include:

Optionally, the news acquisition module may include:

For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.

The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.

With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.

Fig. 5 is a block diagram illustrating an apparatus 900 for news recommendation according to an example embodiment. For example, the apparatus 900 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, an exercise device, a personal digital assistant, and the like.

Referring to fig. 5, the apparatus 900 may include one or more of the following components: processing component 902, memory 904, power component 906, multimedia component 908, audio component 910, input/output (I/O) interface 912, sensor component 914, and communication component 916.

The processing component 902 generally controls overall operation of the device 900, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. Processing element 902 may include one or more processors 920 to execute instructions to perform all or a portion of the steps of the methods described above. Further, processing component 902 can include one or more modules that facilitate interaction between processing component 902 and other components. For example, the processing component 902 can include a multimedia module to facilitate interaction between the multimedia component 908 and the processing component 902.

The memory 904 is configured to store various types of data to support operation at the device 900. Examples of such data include instructions for any application or method operating on device 900, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 904 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.

The power supply component 906 provides power to the various components of the device 900. The power components 906 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 900.

The multimedia component 908 comprises a screen providing an output interface between the device 900 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide motion action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 908 includes a front facing camera and/or a rear facing camera. The front-facing camera and/or the rear-facing camera may receive external multimedia data when the device 900 is in an operating mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

The audio component 910 is configured to output and/or input audio signals. For example, audio component 910 includes a Microphone (MIC) configured to receive external audio signals when apparatus 900 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in the memory 904 or transmitted via the communication component 916. In some embodiments, audio component 910 also includes a speaker for outputting audio signals.

I/O interface 912 provides an interface between processing component 902 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.

The sensor component 914 includes one or more sensors for providing status assessment of various aspects of the apparatus 900. For example, the sensor assembly 914 may detect an open/closed state of the device 900, the relative positioning of the components, such as a display and keypad of the apparatus 900, the sensor assembly 914 may also detect a change in the position of the apparatus 900 or a component of the apparatus 900, the presence or absence of user contact with the apparatus 900, orientation or acceleration/deceleration of the apparatus 900, and a change in the temperature of the apparatus 900. The sensor assembly 914 may include a proximity sensor configured to detect the presence of a nearby object in the absence of any physical contact. The sensor assembly 914 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 914 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 916 is configured to facilitate communications between the apparatus 900 and other devices in a wired or wireless manner. The apparatus 900 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 916 receives a broadcast signal or broadcast associated information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communications component 916 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

In an exemplary embodiment, the apparatus 900 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.

In an exemplary embodiment, a non-transitory computer readable storage medium comprising instructions, such as the memory 904 comprising instructions, executable by the processor 920 of the apparatus 900 to perform the above-described method is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

A non-transitory computer readable storage medium, instructions in which, when executed by a processor of a smart terminal, enable the smart terminal to perform a news recommendation method, the method comprising: acquiring current input content of a current user; when the current input content meets preset news related conditions, acquiring target news corresponding to the main news keywords, or acquiring the main news keywords and the target news corresponding to the related keywords; the main news keywords are matched with the current input content, the main news keywords are obtained according to historical input behavior data of a plurality of users, and the related keywords correspond to the main news keywords; and recommending the target news to the current user.

Fig. 6 is a schematic diagram of a server in some embodiments of the invention. The server 1900 may vary widely by configuration or performance and may include one or more Central Processing Units (CPUs) 1922 (e.g., one or more processors) and memory 1932, one or more storage media 1930 (e.g., one or more mass storage devices) storing applications 1942 or data 1944. Memory 1932 and storage medium 1930 can be, among other things, transient or persistent storage. The program stored in the storage medium 1930 may include one or more modules (not shown), each of which may include a series of instructions operating on a server. Still further, a central processor 1922 may be provided in communication with the storage medium 1930 to execute a series of instruction operations in the storage medium 1930 on the server 1900.

The server 1900 may also include one or more power supplies 1926, one or more wired or wireless network interfaces 1950, one or more input-output interfaces 1958, one or more keyboards 1956, and/or one or more operating systems 1941, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is only limited by the appended claims

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

The news recommending method, the news recommending device and the device for recommending the news provided by the invention are introduced in detail, specific examples are applied in the text to explain the principle and the implementation mode of the invention, and the description of the examples is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.

Claims

1. A news recommendation method, comprising:

acquiring current input content of a current user;

recommending the target news to the current user;

wherein, the current input content accords with preset news related conditions, including:

the current input content is matched with the main news keywords; or

2. The method of claim 1, wherein the news primary keywords are mined by:

3. The method of claim 2, wherein the vocabulary attributes comprise: at least one of word frequency and number of users.

4. The method of claim 3, wherein the preset attribute condition comprises:

5. The method according to any one of claims 2 to 4, further comprising:

6. The method of claim 5, wherein the step of noise filtering the news primary keyword by pseudo-correlation feedback comprises:

7. The method according to any one of claims 1 to 4, wherein the related keywords corresponding to the main news keywords are mined by:

8. The method according to any one of claims 1 to 4, wherein the step of obtaining the target news corresponding to the main news keyword comprises:

9. A news recommender, comprising:

the news recommending module is used for recommending the target news to the current user;

the current input content is matched with the main news keywords; or

10. The apparatus of claim 9, further comprising: the first mining module is used for mining main news keywords;

the first excavation module includes:

11. The apparatus of claim 10, wherein the vocabulary attributes comprise: at least one of word frequency and number of users.

12. The apparatus of claim 10, wherein the preset attribute condition comprises:

13. The apparatus of any of claims 10 to 12, further comprising:

14. The apparatus of claim 13, wherein the noise filtering module comprises:

15. The apparatus of any of claims 9 to 12, further comprising: the second mining module is used for mining related keywords corresponding to the main news keywords;

the second excavation module includes:

16. The apparatus according to any one of claims 9 to 12, wherein the news gathering module comprises:

17. An apparatus for news recommendation comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, the one or more programs comprising instructions for:

acquiring current input content of a current user;

recommending the target news to the current user;

the current input content is matched with the main news keywords; or

18. The apparatus of claim 17, wherein the apparatus is also configured to execute the one or more programs by one or more processors includes instructions for:

19. The apparatus of claim 18, wherein the vocabulary attributes comprise: at least one of word frequency and number of users.

20. The apparatus of claim 18, wherein the preset attribute condition comprises:

21. The apparatus of any of claims 18-20, wherein the apparatus is further configured to execute the one or more programs by one or more processors includes instructions for:

22. The apparatus of claim 21, wherein the noise filtering the news primary keyword by pseudo-correlation feedback comprises:

23. The apparatus of any of claims 17-20, wherein the apparatus is further configured to execute the one or more programs by one or more processors includes instructions for:

24. The apparatus according to any one of claims 17 to 20, wherein the obtaining of the target news corresponding to the main news keyword includes:

25. One or more machine readable media having instructions stored thereon that, when executed by one or more processors, cause an apparatus to perform the method of one or more of claims 1-8.