CN117194794B - Information recommendation method and device, computer equipment and computer storage medium - Google Patents
Information recommendation method and device, computer equipment and computer storage medium Download PDFInfo
- Publication number
- CN117194794B CN117194794B CN202311216224.1A CN202311216224A CN117194794B CN 117194794 B CN117194794 B CN 117194794B CN 202311216224 A CN202311216224 A CN 202311216224A CN 117194794 B CN117194794 B CN 117194794B
- Authority
- CN
- China
- Prior art keywords
- information
- user
- content
- confidence
- recommended content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 230000003044 adaptive effect Effects 0.000 claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims description 13
- 230000000875 corresponding effect Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 8
- 230000002596 correlated effect Effects 0.000 claims description 6
- 238000013075 data extraction Methods 0.000 claims description 6
- 230000006978 adaptation Effects 0.000 claims description 5
- 238000009499 grossing Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 230000010365 information processing Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 abstract description 2
- 210000001508 eye Anatomy 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 210000005252 bulbus oculi Anatomy 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 210000000887 face Anatomy 0.000 description 1
- 230000008921 facial expression Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Abstract
The invention discloses an information recommendation method, an information recommendation device, computer equipment and a computer storage medium, wherein the information recommendation method comprises the following steps: acquiring attribute information of a user; acquiring current scene information; corresponding weights are distributed for the attribute information and the current scene information of the user, and weighted average is carried out on the attribute information and the current scene information of the user, so that comprehensive information is obtained; extracting matched data content from the database according to the comprehensive information; and extracting an influence factor according to the current scene information, converting the matched data content into an adaptive format based on the influence factor, and providing the converted data for a user. The method and the device realize extraction of the content of the matching data based on the acquired attribute information and scene information, can acquire the current required information of the user more accurately, and provide good user experience.
Description
Technical Field
The invention belongs to the field of data information processing, and particularly relates to an information recommendation method, an information recommendation device, computer equipment and a computer storage medium.
Background
With the rapid development of computer, network and multimedia technologies, users acquire needed data information more and more conveniently at any time and any place through intelligent terminals such as PCs, iPads, mobile phones and wearable devices. In the process of using the intelligent terminal, after the intelligent terminal is networked, the user can search and obtain the required data information by inputting the search keyword, and can also inquire and obtain the data stored in advance locally in the intelligent terminal, and the obtained data information is displayed to the user, so that the subsequent use is convenient.
The mode of acquiring the data information requires the user to input retrieval information, or only can query the data stored locally by the intelligent terminal, the provided data is limited, the information input by the user is excessively depended, the use is inconvenient, and the situation that the retrieval of the fed-back data information is not actually required by the user exists.
In the actual use scene, in order to more intelligently recommend the required information to the user, the contents such as the historical browsing record, the geographic position, the interests and the like of the user are obtained, the obtained contents are analyzed, the preference of the user is determined, and more pertinent data contents are recommended to the user based on the preference, so that the method belongs to a common processing scheme, but a large improvement space is provided in the aspects of the accuracy of the recommended data contents, the form of the recommended contents and the like, and the method is a problem to be solved in the industry.
Disclosure of Invention
The invention aims to: in order to solve the problem that information recommended to a user is inaccurate when the user uses an intelligent terminal, the invention provides an information recommendation method, an information recommendation device, computer equipment and a computer storage medium.
The technical scheme is as follows: an information recommendation method, comprising the steps of:
step 1: acquiring attribute information of a user;
step 2: acquiring current scene information;
step 3: corresponding weights are distributed for the attribute information and the current scene information of the user, and weighted average is carried out on the attribute information and the current scene information of the user, so that comprehensive information is obtained;
step 4: extracting matched data content from the database according to the comprehensive information;
step 5: and extracting an influence factor according to the current scene information, converting the matched data content into an adaptive format based on the influence factor, and providing the converted data for a user.
Further, the attribute information includes: age, gender, browsing records, information of interest, and time of work and rest.
Further, the scene information includes: the user is in the environment information, the current time information and the user state.
Further, the step 4 specifically includes:
confidence calculation is carried out on the comprehensive information to obtain confidence information;
and extracting the matched data content from the database according to the confidence information.
Further, the confidence calculation is performed on the comprehensive information to obtain confidence information, and the specific operations include:
vectorizing the comprehensive information and the last recommended content to obtain the comprehensive information and the last recommended content in a vector form, and calculating to obtain the matching degree of the comprehensive information and the last recommended content based on the comprehensive information and the recommended content in the vector form;
extracting the browsing time length of the user on the last recommended content, and forming a plurality of records from the comprehensive information, the browsing time length of the user on the last recommended content and the last recommended content;
normalizing the plurality of records, and scoring each normalized record according to the following formula to obtain the relevance score of the record:
wherein S (c, d) represents the relevance score of user c to recommended content d, k is a smoothing constant, I is the number of recommended content for user c, d i Vector representation, t, representing item i of recommended content i Representing user c versus d i Is a browsing duration of (a);
and setting confidence according to the number, weight and recorded relevance scores of the last recommended content, wherein the confidence is expressed as follows: a=1/(i·p+log S (c, d)); wherein, p represents a weight which is positively correlated with the number of the last recommended content; confidence calculation is carried out based on the confidence coefficient and the comprehensive information, and confidence information is obtained:
wherein M is confidence information, Z is comprehensive information, e is a natural constant, and a is confidence;
and extracting matched data content from the database according to the confidence information.
Further, in step 5, the impact factor is one or more of data format, expected duration, and character preference;
when the influence factor is in the data format, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: converting the matched data text content into video based on the influence factors, or converting the matched data text content into audio based on the influence factors, or converting the matched data video content into text based on the influence factors, or converting the matched data audio content into text based on the influence factors;
when the influence factor is the expected duration, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: dividing the matched data content into a plurality of parts according to time length based on the influence factors;
when the influence factor is character preference, the method for converting the matched data content into the adaptation format based on the influence factor comprises the following steps: the matching data content is translated into a preferred avatar presentation based on the impact factors.
The invention discloses an information recommendation system, which comprises:
the attribute information acquisition module is used for acquiring attribute information of a user;
the current scene information acquisition module is used for acquiring current scene information;
the comprehensive information processing module is used for distributing corresponding weights for the attribute information and the current scene information of the user, and carrying out weighted average on the attribute information and the current scene information of the user to obtain comprehensive information;
the data extraction module is used for extracting data content matched with the comprehensive information from the database according to the comprehensive information;
and the data conversion module is used for extracting influence factors according to the current scene information, converting the matched data content into an adaptive format based on the influence factors, and providing the converted data for a user.
Further, the following steps are performed in the data extraction module:
vectorizing the comprehensive information and the last recommended content to obtain the comprehensive information and the last recommended content in a vector form, and calculating to obtain the matching degree of the comprehensive information and the last recommended content based on the comprehensive information and the recommended content in the vector form;
extracting the browsing time length of the user on the last recommended content, and forming a plurality of records from the comprehensive information, the browsing time length of the user on the last recommended content and the last recommended content;
normalizing the plurality of records, and scoring each normalized record according to the following formula to obtain the relevance score of the record:
wherein S (c, d) represents the relevance score of user c to recommended content d, k is a smoothing constant, I is the number of recommended content for user c, d i Vector representation, t, representing item i of recommended content i Representing user c versus d i Is a browsing duration of (a);
and setting confidence according to the number, weight and recorded relevance scores of the last recommended content, wherein the confidence is expressed as follows: a=1/(i·p+log S (c, d)); wherein, p represents a weight which is positively correlated with the number of the last recommended content; confidence calculation is carried out based on the confidence coefficient and the comprehensive information, and confidence information is obtained:
wherein M is confidence information, Z is comprehensive information, e is a natural constant, and a is confidence;
and extracting matched data content from the database according to the confidence information.
The invention discloses a computer device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the steps of an information recommendation method when executing the computer program.
The present invention discloses a computer storage medium storing a program of an information recommendation method, which when executed by at least one processor implements steps of an information recommendation method.
The beneficial effects are that: compared with the prior art, the invention has the following advantages:
the invention provides a perfect data information recommendation scheme, which is used for acquiring first comprehensive information based on attribute information and current scene information of a user; extracting matched data content from a preset database according to the first comprehensive information; converting the data content, and providing the converted data content for a user; the extraction of the content of the matched data is realized based on the acquired attribute information and scene information, so that the current required information of the user can be acquired more accurately, and good user experience is provided.
Drawings
FIG. 1 is a flow chart of an information recommendation method according to the present invention;
fig. 2 is a schematic diagram of an information recommendation system according to the present invention.
Detailed Description
The technical scheme of the invention is further described with reference to the accompanying drawings and the embodiments.
Example 1:
as shown in fig. 1, the present embodiment discloses an information recommendation method, which can be applied to a plurality of subdivision fields, such as e-commerce, recommendation of multimedia data information, recommendation of educational resources, and the like. The method mainly comprises the following steps:
step 1: acquiring attribute information of a user, wherein the attribute information comprises but is not limited to: age, gender, browse records, information of interest, and work and rest time; the specific operation comprises the following steps:
wherein, the age and sex are obtained according to the following steps:
the method comprises the steps of capturing a face of a currently used intelligent terminal through a front camera of the intelligent terminal, matching the face with a pre-constructed user file, acquiring corresponding file information when a matching result exists, and establishing the user file for the face of the currently used intelligent terminal when the matching result does not exist. The profile information includes age and gender. The construction of the user file can be based on the information pre-registered by the user, or a deep neural network face recognition algorithm can be adopted to recognize the face photo in the user album, so that the age and sex corresponding to the face in the face photo are obtained, and the corresponding user file is built for different faces. In the user profile: age, gender, browsing records, information of interest, and time of work and rest.
The browse records of this embodiment include, but are not limited to: page access time, access frequency, UI operation (such as zoom, sliding speed, selected data, click position, etc.), dwell time, association between front and back pages, page theme, shopping record; the residence time is obtained according to the following operation: and when the user browses information, the facial expression of the user is identified through an artificial intelligence algorithm, and the expression can represent the preference of the user to the content displayed on the current page. And carrying out binarization processing on the acquired multi-frame continuous face images, extracting all connected areas in the binarized images for each frame of image, carrying out area detection on all the connected areas, sorting from large to small, selecting connected area combinations with adjacent area differences smaller than a first threshold value as candidate areas, comparing the areas of the candidate areas in the multi-frame, and taking the candidate areas with front-rear area changes exceeding a second threshold value as eye areas, wherein the second threshold value is larger than the first threshold value, so that the positioning of the eye areas in the face images is realized. The gray value of each point in the eye area is calculated, the gray value of each point is differed from the gray values of other adjacent points in the eight adjacent points of the point, the point is connected with the adjacent point with the largest difference, each connecting line is extended to two ends, the point with the most connecting lines intersecting in the eye area is obtained, and the point is taken as the center position of the eyeball. And acquiring face information of the user based on the front camera, and acquiring pose information of the terminal including the position and the pose inclination angle. Based on pose information of the terminal and central position change information of eyeballs of a user, focusing information of eyes on a terminal screen is positioned, further focal content of current focus of the user is determined, and duration of the user focusing on the focal content is recorded.
The attention information may be based on data pre-filled by the user, or may be obtained based on the above-described analysis of the browsing records. If the information of interest is obtained based on the above analysis of the browsing records, the specific obtaining steps include: clustering operation is carried out on page topics in the browsing records, topic distribution conditions of content pages accessed by users are obtained, and information association is carried out on the topic distribution conditions, access frequency and access time, so that attention information of the users is obtained.
Step 2: current context information is obtained including, but not limited to, ambient information, time information, user status. The surrounding environment information comprises, but is not limited to, environment information of a scene in which the user is located, weather information, the number of surrounding people and environment volume; the time information includes, but is not limited to, date information, current time information, user's schedule record, alarm clock content information; user states include stationary, moving. The current scene information can assist in achieving judgment that the user is in an activity state such as outdoor activities, indoor sitting, working/class, leisure/busy and the like. The surrounding environment information can be obtained through a front-facing camera, and the state of the terminal can be detected through a sensor in the terminal. And judging whether the user is in an exercise running state according to the exercise condition of the terminal, further judging according to the time, acquiring weather information in a networking manner, and acquiring environmental volume through a microphone of the terminal. Step 3: the influence degree of the attribute information and the scene information of the user on whether the recommended data information is matched with the actual needs of the user is different, and the influence degree can be dynamically changed in different scenes. In the implementation, the attribute information and the scene information can be presented in a key value pair or vectorization mode, corresponding influence weights are respectively allocated to the attribute information and the scene information, but the weights dynamically change along with the attribute information and the scene information, for example, when the acquired attribute information is small in quantity and low in quality, the allocated weights are relatively small, and the acquired related attribute information is more and more rich along with the continuous updating of the user file, so that the allocated weights are increased. The corresponding assigned weights may also be different in different scenarios. For example, when the acquired scene information characterizes that the user is more focused on using the terminal, a higher weight is allocated to the scene information. Therefore, corresponding weights are allocated to the attribute information and the current scene information of the user, and are recorded as alpha, beta, alpha and beta which dynamically change along with the attribute information and the scene information, and comprehensive information is obtained by means of weighted average of the attribute information and the scene information, and in the embodiment, the initial value of alpha and beta is preferably 0.45, and the sum of the initial value and the initial value can exceed 1. The comprehensive information mainly reflects two parts of contents, namely information related to the user and information of the periphery of the user, wherein the information related to the user can comprise information related to identity attributes, daily-focused field contents and the like, the periphery information can be related to surrounding environments, weather and the like, when data content recommendation is carried out, the two parts of contents are comprehensively considered, the content interested by the user can be recommended in a proper occasion, and user experience is improved.
Step 4: extracting matched data content from the database according to the comprehensive information; the database mentioned in the step is pre-constructed by the service provider, the data in the database is stored according to categories, such as articles, media resources and course resources, and the data in the database is classified according to a plurality of dimensions such as topics, expression forms, duration, difficulty coefficients and the like, so that the multi-mapping storage is realized. The specific operation comprises the following steps:
because the individuation of the data of the single user is strong, the quality of the recommended content obtained by directly carrying out matching based on the comprehensive information is possibly lower, and in order to improve the matching degree of the data content and the user requirement, the confidence calculation is carried out on the comprehensive information, so that the information correction is realized. After the comprehensive information is acquired, confidence calculation is carried out on the comprehensive information, confidence information is acquired, and the confidence information represents the interest degree of the user on the set category data; the specific implementation method comprises the following steps:
s400: acquiring comprehensive information of a lot of users and recommended contents which have been recommended to the users before the time, wherein the embodiment is based on the past recommended contents, and further provides the user with data content matched and screened from a database by combining the comprehensive information of the users, so that the adaptation degree of the provided data and the users is improved; the matched data content extracted from the database is the new recommended data provided after the optimization matching;
s410: vectorizing the comprehensive information and the recommended content which is recommended to the user before the present time to obtain comprehensive information and recommended content in a vector form, calculating the matching degree of the comprehensive information and the recommended content in a subsequent step, and calculating the similarity of the comprehensive information and the recommended content in the vector form in a cosine similarity mode, a Euclidean distance mode and the like to obtain the matching degree of the comprehensive information and the recommended content;
s420: and extracting the time length information of the user for browsing the recommended content, wherein the time length information of the user for browsing the recommended content refers to the time length statistical information of the recommended content which is previously browsed by the user and recommended to the user, and the browsing time length represents the interest degree of the user for the data content.
S430: forming a plurality of data records by using the time length information, the comprehensive information and the matching degree of the recommended content of the user, and deleting the data records with the time length less than a third threshold value from the plurality of data records; a data record may, for example, resemble a line of data in a table, such as "data content browsing duration match".
S440: normalizing the data records, and scoring each record according to the following formula:
wherein S (c, d) represents the relevance score of the user c to the recommended content d, k is a smooth constant, the value range is 0 to 1, I is the number of recommended contents for the user c in the record set, d i Vector representing ith recommended contentRepresentation, t i Representing user c versus d i Is a browsing duration of (a).
S450: confidence information is obtained based on scoring conditions of all users in all record sets, specifically, corresponding weights are distributed to the users based on the quantity of recommended contents of each user in the record sets, the weights are positively correlated with the quantity of the contents, namely, the confidence is set according to the quantity of recommended contents for the user c, the weights and the relevance score of the records: the confidence level indicates the approval degree of the user on the past recommended content to a certain extent, and is related to the quantity and the relativity of the recommended content, and a specific setting formula is as follows: a=1/(i·p+log S (c, d)), where I is the number of recommended contents and p represents a weight.
S460: confidence calculation is achieved based on the confidence coefficient and the comprehensive information, confidence information is obtained, the confidence information can more accurately represent the interest degree of a user in specific content, and matched data content is extracted from the database according to the confidence information.
The confidence calculation process comprises the following steps:
wherein M is confidence information, Z is comprehensive information, e is a natural constant, and a is confidence.
Step 5: and extracting an influence factor according to the current scene information, converting the matched data content into an adaptive format based on the influence factor, and providing the converted data for a user. In the daily use process of the terminal, under different scene conditions, the user generally has different requirements on the display form of the data content information. For example, when the user is in exercise state such as exercise/running, the acceptance of data content in audio format is high; when the time of the user is limited, the method is more suitable for media content with shorter content and shorter time; in silent environments, data content in text format tends to be read; according to different scene information, a plurality of influence factors including at least a data format, a predicted duration and character preference can be set, according to the actual scene, the data content to be recommended is converted into video/audio according to the influence factors, for example, text is converted into video/audio, or vice versa, the text/media is divided into a plurality of parts according to duration, the video/audio is converted into a preferred virtual character description, and the converted data is provided for a user. After the data content is provided for the user, the feedback information of the user on the display content including UI operation is monitored, the update of the user file is realized, and reference information is provided for the follow-up data content adjustment recommendation process.
Example 2:
referring to fig. 2, the present embodiment discloses an information recommendation system, which mainly includes:
the attribute information acquisition module is used for acquiring attribute information of a user;
the current scene information acquisition module is used for acquiring current scene information;
the comprehensive information processing module is used for distributing corresponding weights for the attribute information and the current scene information of the user, and carrying out weighted average on the attribute information and the current scene information of the user to obtain comprehensive information;
the data extraction module is used for extracting data content matched with the comprehensive information from the database according to the comprehensive information;
and the data conversion module is used for extracting influence factors according to the current scene information, converting the matched data content into an adaptive format based on the influence factors, and providing the converted data for a user. The impact factor is one or more of data format, expected duration, and character preference.
The following steps are performed in the data extraction module:
vectorizing the comprehensive information and the last recommended content to obtain the comprehensive information and the last recommended content in a vector form, and calculating to obtain the matching degree of the comprehensive information and the last recommended content based on the comprehensive information and the recommended content in the vector form;
extracting the browsing time length of the user on the last recommended content, and forming a plurality of records from the comprehensive information, the browsing time length of the user on the last recommended content and the last recommended content;
normalizing the plurality of records, and scoring each normalized record according to the following formula to obtain the relevance score of the record:
wherein S (c, d) represents the relevance score of user c to recommended content d, k is a smoothing constant, I is the number of recommended content for user c, d i Vector representation, t, representing item i of recommended content i Representing user c versus d i Is a browsing duration of (a);
and setting confidence according to the number, weight and recorded relevance scores of the last recommended content, wherein the confidence is expressed as follows: a=1/(i·p+log S (c, d)); wherein, p represents a weight which is positively correlated with the number of the last recommended content; confidence calculation is carried out based on the confidence coefficient and the comprehensive information, and confidence information is obtained:
wherein M is confidence information, Z is comprehensive information, e is a natural constant, and a is confidence;
and extracting matched data content from the database according to the confidence information.
The following steps are performed in the data conversion module:
when the influence factor is in the data format, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: converting the matched data text content into video based on the influence factors, or converting the matched data text content into audio based on the influence factors, or converting the matched data video content into text based on the influence factors, or converting the matched data audio content into text based on the influence factors;
when the influence factor is the expected duration, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: dividing the matched data content into a plurality of parts according to time length based on the influence factors;
when the influence factor is character preference, the method for converting the matched data content into the adaptation format based on the influence factor comprises the following steps: the matching data content is translated into a preferred avatar presentation based on the impact factors.
Example 3:
the embodiment discloses a computer device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program to realize the steps disclosed in any one of the embodiments.
Example 4:
the present embodiment discloses a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps disclosed in any of the embodiments described above.
Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, or the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples merely represent a few embodiments of the present application, which are described in more detail and are not to be construed as limiting the scope of the invention. It should be noted that it would be apparent to those skilled in the art that various modifications and improvements could be made without departing from the spirit of the present application, which would be within the scope of the present application. Accordingly, the scope of protection of the present application is to be determined by the claims appended hereto.
Claims (6)
1. An information recommendation method is characterized in that: the method comprises the following steps:
step 1: acquiring attribute information of a user;
step 2: acquiring current scene information;
step 3: corresponding weights are distributed for the attribute information and the current scene information of the user, and weighted average is carried out on the attribute information and the current scene information of the user, so that comprehensive information is obtained;
step 4: extracting matched data content from the database according to the comprehensive information;
step 5: extracting an influence factor according to the current scene information, converting the matched data content into an adaptive format based on the influence factor, and providing the converted data for a user;
the step 4 specifically includes:
confidence calculation is carried out on the comprehensive information to obtain confidence information;
extracting matched data content from the database according to the confidence information;
the confidence calculation is carried out on the comprehensive information to obtain the confidence information, and the specific operations comprise:
vectorizing the comprehensive information and the last recommended content to obtain the comprehensive information and the last recommended content in a vector form, and calculating to obtain the matching degree of the comprehensive information and the last recommended content based on the comprehensive information and the recommended content in the vector form;
extracting the browsing time length of the user on the last recommended content, and forming a plurality of records from the comprehensive information, the browsing time length of the user on the last recommended content and the last recommended content;
normalizing the plurality of records, and scoring each normalized record according to the following formula to obtain the relevance score of the record:
wherein S (c, d) represents the relevance score of user c to recommended content d, k is a smoothing constant, I is the number of recommended content for user c, d i Vector representation, t, representing item i of recommended content i Representing user c versus d i Is a browsing duration of (a);
and setting confidence according to the number, weight and recorded relevance scores of the last recommended content, wherein the confidence is expressed as follows: a=1/(i·p+logs (c, d)); wherein, p represents a weight which is positively correlated with the number of the last recommended content; confidence calculation is carried out based on the confidence coefficient and the comprehensive information, and confidence information is obtained:
wherein M is confidence information, Z is comprehensive information, e is a natural constant, and a is confidence;
extracting matched data content from the database according to the confidence information;
in step 5, the impact factor is one or more of data format, expected duration and role preference;
when the influence factor is in the data format, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: converting the matched data text content into video based on the influence factors, or converting the matched data text content into audio based on the influence factors, or converting the matched data video content into text based on the influence factors, or converting the matched data audio content into text based on the influence factors;
when the influence factor is the expected duration, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: dividing the matched data content into a plurality of parts according to time length based on the influence factors;
when the influence factor is character preference, the method for converting the matched data content into the adaptation format based on the influence factor comprises the following steps: the matching data content is translated into a preferred avatar presentation based on the impact factors.
2. An information recommendation method according to claim 1, characterized in that: the attribute information includes: age, gender, browsing records, information of interest, and time of work and rest.
3. An information recommendation method according to claim 2, characterized in that: the scene information includes: the user is in the environment information, the current time information and the user state.
4. An information recommendation system, characterized in that: comprising the following steps:
the attribute information acquisition module is used for acquiring attribute information of a user;
the current scene information acquisition module is used for acquiring current scene information;
the comprehensive information processing module is used for distributing corresponding weights for the attribute information and the current scene information of the user, and carrying out weighted average on the attribute information and the current scene information of the user to obtain comprehensive information;
the data extraction module is used for extracting data content matched with the comprehensive information from the database according to the comprehensive information;
the data conversion module is used for extracting influence factors according to the current scene information, converting matched data content into an adaptive format based on the influence factors and providing the converted data for a user;
the following steps are performed in the data extraction module:
vectorizing the comprehensive information and the last recommended content to obtain the comprehensive information and the last recommended content in a vector form, and calculating to obtain the matching degree of the comprehensive information and the last recommended content based on the comprehensive information and the recommended content in the vector form;
extracting the browsing time length of the user on the last recommended content, and forming a plurality of records from the comprehensive information, the browsing time length of the user on the last recommended content and the last recommended content;
normalizing the plurality of records, and scoring each normalized record according to the following formula to obtain the relevance score of the record:
wherein S (c, d) represents the relevance score of user c to recommended content d, k is a smoothing constant, I is the number of recommended content for user c, d i Vector representation, t, representing item i of recommended content i Representing user c versus d i Is a browsing duration of (a);
and setting confidence according to the number, weight and recorded relevance scores of the last recommended content, wherein the confidence is expressed as follows: a=1/(i·p+logs (c, d)); wherein, p represents a weight which is positively correlated with the number of the last recommended content; confidence calculation is carried out based on the confidence coefficient and the comprehensive information, and confidence information is obtained:
wherein M is confidence information, Z is comprehensive information, e is a natural constant, and a is confidence;
extracting matched data content from the database according to the confidence information;
the influence factors are one or more of data format, expected duration and role preference;
when the influence factor is in the data format, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: converting the matched data text content into video based on the influence factors, or converting the matched data text content into audio based on the influence factors, or converting the matched data video content into text based on the influence factors, or converting the matched data audio content into text based on the influence factors;
when the influence factor is the expected duration, the method for converting the matched data content into the adaptive format based on the influence factor comprises the following steps: dividing the matched data content into a plurality of parts according to time length based on the influence factors;
when the influence factor is character preference, the method for converting the matched data content into the adaptation format based on the influence factor comprises the following steps: the matching data content is translated into a preferred avatar presentation based on the impact factors.
5. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of an information recommendation method according to any one of claims 1 to 3 when the computer program is executed.
6. A computer storage medium, characterized in that the storage medium stores a program of an information recommendation method, which when executed by at least one processor implements the steps of an information recommendation method according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311216224.1A CN117194794B (en) | 2023-09-20 | 2023-09-20 | Information recommendation method and device, computer equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311216224.1A CN117194794B (en) | 2023-09-20 | 2023-09-20 | Information recommendation method and device, computer equipment and computer storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117194794A CN117194794A (en) | 2023-12-08 |
| CN117194794B true CN117194794B (en) | 2024-03-26 |
Family
ID=88983256
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311216224.1A Active CN117194794B (en) | 2023-09-20 | 2023-09-20 | Information recommendation method and device, computer equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117194794B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016150170A1 (en) * | 2015-03-25 | 2016-09-29 | 百度在线网络技术（北京）有限公司 | Recommendation method, device and equipment and computer storage medium |
| CN107436893A (en) * | 2016-05-26 | 2017-12-05 | 北京搜狗科技发展有限公司 | A kind of webpage recommending method and device |
| CN109121007A (en) * | 2018-09-18 | 2019-01-01 | 深圳市酷开网络科技有限公司 | Movie and television contents recommended method, smart television and system based on plurality of human faces identification |
| CN110059250A (en) * | 2019-04-18 | 2019-07-26 | 广东小天才科技有限公司 | Information recommendation method, device, equipment and storage medium |
| CN113325767A (en) * | 2021-05-27 | 2021-08-31 | 深圳Tcl新技术有限公司 | Scene recommendation method and device, storage medium and electronic equipment |
| CN114048335A (en) * | 2021-11-09 | 2022-02-15 | 江苏科技大学 | Knowledge base-based user interaction method and device |
| CN114579858A (en) * | 2022-03-03 | 2022-06-03 | 平安科技（深圳）有限公司 | Content recommendation method and device, electronic equipment and storage medium |
| CN114595372A (en) * | 2020-12-03 | 2022-06-07 | 青岛海尔智能技术研发有限公司 | Scene recommendation method and device, computer equipment and storage medium |
| CN115186177A (en) * | 2022-06-15 | 2022-10-14 | 平安银行股份有限公司 | Social information recommendation method and device, computer equipment and storage medium |
| CN116319138A (en) * | 2023-03-24 | 2023-06-23 | 广东好太太科技集团股份有限公司 | Intelligent device scene recommendation method and device, storage medium and computer device |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111291266B (en) * | 2020-02-13 | 2023-03-21 | 深圳市雅阅科技有限公司 | Artificial intelligence based recommendation method and device, electronic equipment and storage medium |
-
2023
- 2023-09-20 CN CN202311216224.1A patent/CN117194794B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016150170A1 (en) * | 2015-03-25 | 2016-09-29 | 百度在线网络技术（北京）有限公司 | Recommendation method, device and equipment and computer storage medium |
| CN107436893A (en) * | 2016-05-26 | 2017-12-05 | 北京搜狗科技发展有限公司 | A kind of webpage recommending method and device |
| CN109121007A (en) * | 2018-09-18 | 2019-01-01 | 深圳市酷开网络科技有限公司 | Movie and television contents recommended method, smart television and system based on plurality of human faces identification |
| CN110059250A (en) * | 2019-04-18 | 2019-07-26 | 广东小天才科技有限公司 | Information recommendation method, device, equipment and storage medium |
| CN114595372A (en) * | 2020-12-03 | 2022-06-07 | 青岛海尔智能技术研发有限公司 | Scene recommendation method and device, computer equipment and storage medium |
| CN113325767A (en) * | 2021-05-27 | 2021-08-31 | 深圳Tcl新技术有限公司 | Scene recommendation method and device, storage medium and electronic equipment |
| CN114048335A (en) * | 2021-11-09 | 2022-02-15 | 江苏科技大学 | Knowledge base-based user interaction method and device |
| CN114579858A (en) * | 2022-03-03 | 2022-06-03 | 平安科技（深圳）有限公司 | Content recommendation method and device, electronic equipment and storage medium |
| CN115186177A (en) * | 2022-06-15 | 2022-10-14 | 平安银行股份有限公司 | Social information recommendation method and device, computer equipment and storage medium |
| CN116319138A (en) * | 2023-03-24 | 2023-06-23 | 广东好太太科技集团股份有限公司 | Intelligent device scene recommendation method and device, storage medium and computer device |
Non-Patent Citations (2)
| Title |
|---|
| 基于长短期偏好的自适应融合推荐算法;周倩 等;《江苏科技大学自然学报（自然科学版）》;第37卷(第4期);57-64 * |
| 融合知识图谱与用户病情画像的在线医疗社区场景化信息推荐研究;翟姗姗 等;《情报科学》;第39卷(第5期);97-105 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117194794A (en) | 2023-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111444428B (en) | Information recommendation method and device based on artificial intelligence, electronic equipment and storage medium | |
| CN109800352B (en) | Method, system and terminal device for pushing information based on clipboard | |
| US7684651B2 (en) | Image-based face search | |
| CN107590224B (en) | Big data based user preference analysis method and device | |
| CN113158023B (en) | Public digital life accurate classification service method based on mixed recommendation algorithm | |
| CN111310019A (en) | Information recommendation method, information processing method, system and equipment | |
| CN112052387B (en) | Content recommendation method, device and computer readable storage medium | |
| CN109582869B (en) | Data processing method and device and data processing device | |
| CN108959323B (en) | Video classification method and device | |
| KR20170131924A (en) | Method, apparatus and computer program for searching image | |
| CN113806588B (en) | Method and device for searching video | |
| CN112364204A (en) | Video searching method and device, computer equipment and storage medium | |
| CN111512299A (en) | Method for content search and electronic device thereof | |
| CN112069326A (en) | Knowledge graph construction method and device, electronic equipment and storage medium | |
| CN110110218B (en) | Identity association method and terminal | |
| JP5611155B2 (en) | Content tagging program, server and terminal | |
| CN111223014B (en) | Method and system for online generation of subdivision scene teaching courses from a large number of subdivision teaching contents | |
| CN117194794B (en) | Information recommendation method and device, computer equipment and computer storage medium | |
| CN111797765B (en) | Image processing method, device, server and storage medium | |
| CN113902526A (en) | Artificial intelligence based product recommendation method and device, computer equipment and medium | |
| CN111222011B (en) | Video vector determining method and device | |
| CN115935049A (en) | Recommendation processing method and device based on artificial intelligence and electronic equipment | |
| CN110351183B (en) | Resource collection method and device in instant messaging | |
| TW202004516A (en) | Optimization method for searching exclusive personalized pictures | |
| US20220164090A1 (en) | Abstract generation method and apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |