CN114722290A - Trust-relationship-fused ranking learning POI recommendation algorithm - Google Patents
Trust-relationship-fused ranking learning POI recommendation algorithm Download PDFInfo
- Publication number
- CN114722290A CN114722290A CN202210395703.3A CN202210395703A CN114722290A CN 114722290 A CN114722290 A CN 114722290A CN 202210395703 A CN202210395703 A CN 202210395703A CN 114722290 A CN114722290 A CN 114722290A
- Authority
- CN
- China
- Prior art keywords
- path
- user
- learning
- users
- fused
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9538—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the field of data processing, and discloses a trust relationship fused ranking learning POI recommendation algorithm, which comprises the following steps: step 1: collecting user footprint information, recording and uploading characteristics which can be used as reference bases; step 2: analyzing the characteristics, extracting key words and screening out interference words; step 3: generating a dedicated corresponding label according to each characteristic; step 4: calculating the access times of the corresponding label features, and sequencing according to the access times; step 5: the method comprises the steps of collecting historical social behaviors of users, associating social networks of the users, capturing social relations, and taking the associated users as recommenders. The method and the system can cut in the social network of the user, enable the user or other users to serve as recommendation points for guidance, drive the enthusiasm of the user, improve the receiving degree of the recommended items, can accurately grasp the preference requirements of the user, improve the use experience of the user, and provide in-place personalized recommendation for the user.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a trust relationship fused ranking learning POI recommendation algorithm.
Background
The search results are ranked by using a machine learning technology, the search results are continuously developed as a popular research field, a traditional information retrieval model ranks the relevance of examination and documents, the considered factors are not many, the ranking formula is mainly manually fitted by using several factors, namely word frequency, inverse document frequency and document length, and with the development of a search engine, more and more factors are needed to be considered for ranking a certain webpage or a certain project, and the factors are more and more complex;
the existing recommendation algorithm is not enough in place for personalized recommendation of a user, so that the interest of the user is difficult to accurately grasp, and the social network of the user cannot be used as an entry point, so that other users in the social network are difficult to drive the enthusiasm of the user as guidance, and the preference demand of the user is difficult to guess.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects in the prior art, the invention provides a trust relationship fused ranking learning POI recommendation algorithm, which can effectively solve the problems that personalized recommendation of a user is not in place, the interest of the user is difficult to accurately grasp, other users in a social network are difficult to drive as guidance, the enthusiasm of the user is difficult to drive, and the preference requirement of the user is difficult to measure in the prior art.
(II) technical scheme
In order to achieve the above objects, the present invention is achieved by the following technical solutions,
the invention discloses a trust relationship fused ranking learning POI recommendation algorithm, which comprises the following steps:
step 1: collecting user footprint information, recording and uploading characteristics which can be used as reference bases;
step 2: analyzing the characteristics, extracting key words and screening out interference words;
step 3: generating a corresponding exclusive label according to each characteristic;
step 4: calculating the access times of the corresponding label features, and sequencing according to the access times;
step 5: collecting historical social behaviors of users, associating the social networks of the users, capturing social relations, and taking the associated users as recommenders;
step 6: collecting target users with high trust degrees on the associated users;
step 7: generating a feasible arrival path according to the corresponding label;
step 8: analyzing the cost of a feasible arriving path, taking the arriving path with the lowest cost as the optimal priority, and collecting and analyzing interference factors as reference data;
step 9: recommending the priority path, and sending the priority path to a project recommendation column of a target user in real time;
step 10: and when the target user accesses the network and triggers the item under the corresponding label, recommending the item and mentioning the associated user information.
Further, the feature recording manner in Step1, including the feature selection algorithm based on statistical learning, is divided into based on correlation measurement and feature space representation;
and (4) utilizing the feature words obtained by feature selection to segment the extracted words and sentences, discarding the name core words and keeping the rest.
Further, the evaluation index of the feature keyword in step2 includes: coverage and accuracy;
wherein the coverage refers to the proportion of the sample number predictable by the algorithm in the whole test sample set;
the algorithm accuracy rate refers to the proportion of the correct samples in the whole test sample set can be predicted.
Further, the attribute of the attribute corresponding to the tag in step3 includes: an independent storage location, a unique trigger condition, and an encrypted transmission channel.
Further, the sorting in Step4 is performed by sequentially sorting the tags with high access times as data to be preferentially associated, and then associating one by one.
Further, the gathering of the path of the target user in Step6 includes: interfacing the social network, the platform footprint record, and the software project registration record over a wireless network.
Further, the reference manner of the associated user in the step10 includes: instant popup prompts, in-project message prompts and mobile device editing message prompts.
Further, the process of collecting and analyzing the interference factors in Step8 includes the following steps:
s1: when the path analysis is triggered, the server network is linked;
s2: collecting current path information participating in analysis and collecting path data;
s3: recording real-time path information and calling historical path data;
s4: referring to the path selection information of the associated user, sorting by how many times the user has visited;
s5: the user edits and sets the future time period of the current path;
s6: predicting the path influence factors of the editing time period by taking the real-time path information and the historical path data as reference data;
s7: comprehensively sequencing according to the cost consumed by the influence factors;
s8: displaying the sequencing result, providing an optimal path and a secondary path, screening out the residual paths, and automatically confirming the optimal path in a default state;
s9: and recording the selection result and generating a work report.
Further, the attributes of the path information in step S2 include: current road conditions, current weather conditions, and vehicle restrictions.
Further, the attributes of the job report in step S9 include: recording time, recording location, generated path information, and path selection node.
(III) advantageous effects
Compared with the known public technology, the technical scheme provided by the invention has the following beneficial effects,
1. the method and the system can be switched in aiming at the social network of the user, enable the user or other users to serve as recommendation points for guidance, drive the enthusiasm of the user, improve the receiving degree of the recommended items, can accurately grasp the preference requirements of the user, can improve the use experience of the user, can drive the development of related items, can provide in-place personalized recommendation for the user, and have high practicability.
2. The invention further stimulates the enthusiasm of the user by screening the project recommendation paths, helps the user to save cost, can use historical data as reference data, improves the data processing speed, gradually improves the recommendation paths in the long-term use process, and is further beneficial to developing more users participating in recommendation.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
FIG. 1 is a schematic flow diagram of a trust relationship fused ranking learning POI recommendation algorithm;
FIG. 2 is a schematic flow chart of a process of collecting and analyzing interference factors according to the present invention;
FIG. 3 is a schematic diagram illustrating an architecture of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The present invention will be further described with reference to the following examples.
Example 1
The trust relationship fused ranking learning POI recommendation algorithm of the embodiment, as shown in FIGS. 1 and 2, includes the following steps:
step 1: collecting user footprint information, recording and uploading characteristics which can be used as reference bases;
step 2: analyzing the characteristics, extracting key words and screening out interference words;
step 3: generating a corresponding exclusive label according to each characteristic;
step 4: calculating the access times of the corresponding label features, and sequencing according to the access times;
step 5: collecting historical social behaviors of users, associating the social networks of the users, capturing social relations, and taking the associated users as recommenders;
step 6: collecting target users with high trust degrees on the associated users;
step 7: generating a feasible arrival path according to the corresponding label;
step 8: analyzing the cost of a feasible arriving path, taking the arriving path with the lowest cost as the optimal priority, and collecting and analyzing interference factors as reference data;
step 9: recommending the priority path, and sending the priority path to a project recommendation column of a target user in real time;
step 10: and when the target user accesses the network and triggers the item under the corresponding label, recommending the item and referring to the associated user information.
As shown in fig. 1, the manner of feature recording in Step1, including the feature selection algorithm based on statistical learning, is divided into correlation metric and feature space representation;
and (4) utilizing the feature words obtained by feature selection to segment the extracted words and sentences, discarding the name core words and keeping the rest.
As shown in fig. 1, the evaluation index of the feature keyword in step2 includes: coverage and accuracy;
wherein the coverage refers to the proportion of the sample number predictable by the algorithm in the whole test sample set;
the algorithm accuracy rate refers to the proportion of the correct samples in the whole test sample set can be predicted.
As shown in fig. 1, the attribute of the attribute-specific corresponding tag in step3 includes: an independent storage location, a unique trigger condition, and an encrypted transmission channel.
As shown in fig. 1, the sorting in Step4 is performed by sequentially sorting the data, which are associated with priority, by the tags having the high access frequency.
As shown in fig. 1, the gathering of the target user path in Step6 includes: interfacing the social network, the platform footprint record, and the software project registration record over a wireless network.
As shown in fig. 1, the reference manner of the associated user in the step10 includes: instant popup prompts, in-project message prompts, and mobile device edit message prompts.
Through the setting, the social network of the user can be cut in, the user or other users can be guided as recommendation points, the enthusiasm of the user is driven, the receiving degree of the recommendation items is improved, the preference requirements of the user can be mastered accurately, the use experience of the user can be improved, the development of related items can be driven, the in-place personalized recommendation can be provided for the user, and the practicability is high.
Example 2
In other aspects, this embodiment further provides a process for collecting and analyzing the interference factors, as shown in fig. 2, the process for collecting and analyzing the interference factors in Step8 includes the following steps:
s1: when the path analysis is triggered, linking into a server network;
s2: collecting current path information participating in analysis and collecting path data;
s3: recording real-time path information and calling historical path data;
s4: referring to the path selection information of the associated user, sorting by how many times the user has visited;
s5: the user edits and sets the future time period of the current path;
s6: predicting the path influence factors of the editing time period by taking the real-time path information and the historical path data as reference data;
s7: comprehensively sorting according to the cost consumed by the influence factors;
s8: displaying the sequencing result, providing an optimal path and a secondary path, screening out the residual paths, and automatically confirming the optimal path in a default state;
s9: and recording the selection result and generating a work report.
As shown in fig. 2, the attributes of the path information in step S2 include: current road conditions, current weather conditions, and vehicle restrictions.
As shown in fig. 2, the attributes of the job report in step S9 include: recording time, recording location, generated path information, and path selection result.
Through the arrangement, the enthusiasm of the user is further stimulated through screening of the project recommendation paths, the user is helped to save cost, historical data can be used as reference data, the data processing speed is increased, the recommendation paths are gradually improved in the long-term use process, and further more users participating in recommendation can be developed.
Example 3
In this example, the trust recommendation algorithm based on the memory is based on the extended breadth-first search to obtain a recommendation list of the user, the recommendation user is judged according to the path distance when being searched, and then the trust value between the target user and the recommendation user is used as the weight so as to predict the score of the target user on the target item. And when the trust value between the users is high, determining the local trust according to the historical behaviors of the users, and further determining the global trust between the users without direct contact.
In conclusion, the method and the device can cut in the social network of the user, enable the user or other users to serve as recommendation points for guidance, drive the enthusiasm of the user, improve the receptivity of the recommended items, can accurately grasp the preference requirements of the user, can improve the use experience of the user, can drive the development of related items, can provide in-place personalized recommendation for the user, and have high practicability;
by screening the project recommendation paths, the enthusiasm of the user is further stimulated, the user is helped to save cost, historical data can be used as reference data, the data processing speed is increased, the recommendation paths are gradually improved in the long-term use process, and further more users participating in recommendation can be developed.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention; such modifications and substitutions do not depart from the spirit and scope of the corresponding technical solutions.
Claims (10)
1. A trust relationship fused ranking learning POI recommendation algorithm is characterized by comprising the following steps:
step 1: collecting user footprint information, recording and uploading characteristics which can be used as reference bases;
step 2: analyzing the characteristics, extracting key words and screening out interference words;
step 3: generating a corresponding exclusive label according to each characteristic;
step 4: calculating the access times of the corresponding label features, and sequencing according to the access times;
step 5: collecting historical social behaviors of users, associating the social networks of the users, capturing social relations, and taking the associated users as recommenders;
step 6: collecting target users with high trust degrees on the associated users;
step 7: generating a feasible arrival path according to the corresponding label;
step 8: analyzing the cost of a feasible arriving path, taking the arriving path with the lowest cost as the optimal priority, and collecting and analyzing interference factors as reference data;
step 9: recommending the priority path, and sending the priority path to a project recommendation column of a target user in real time;
step 10: and when the target user accesses the network and triggers the item under the corresponding label, recommending the item and referring to the associated user information.
2. The POI recommendation algorithm of rank learning fused with trust relationship as claimed in claim 1, wherein the feature recording manner in Step1 includes feature selection algorithm based on statistical learning, which is divided into relevance metric and feature space representation;
and (4) utilizing the feature words obtained by feature selection to segment the extracted words and sentences, discarding the name core words and keeping the rest.
3. The POI recommendation algorithm for ranked learning fused with trust relationship according to claim 1, wherein the evaluation index of the feature keyword in the step2 comprises: coverage and accuracy;
wherein the coverage refers to the proportion of the sample number predictable by the algorithm in the whole test sample set;
the algorithm accuracy rate refers to the proportion of the correct samples in the whole test sample set can be predicted.
4. The POI recommendation algorithm for ranked learning fused with trust relationship according to claim 1, wherein the attribute of the tag specifically corresponding to the step3 comprises: an independent storage location, a unique trigger condition, and an encrypted transmission channel.
5. The POI recommendation algorithm for ranked learning fused with trust relationship according to claim 1, wherein the ranking in Step4 is performed by taking a tag with a high number of accesses as data to participate in association preferentially, and performing association one by one after being arranged in sequence.
6. The POI recommendation algorithm for learning-by-ranking with fusion trust relationship of claim 1, wherein the Step6 of gathering the path of the target user comprises: interfacing the social network, the platform footprint record, and the software project registration record over a wireless network.
7. The POI recommendation algorithm for learning ranked POIs fused with trust relationship as claimed in claim 1, wherein the reference manner of the associated user in the step10 comprises: instant popup prompts, in-project message prompts, and mobile device edit message prompts.
8. The POI recommendation algorithm for learning by ranking fused with trust relationship according to claim 1, wherein the process of collecting and analyzing the interference factors in Step8 comprises the following steps:
s1: when the path analysis is triggered, linking into a server network;
s2: collecting current path information participating in analysis and collecting path data;
s3: recording real-time path information and calling historical path data;
s4: referring to the path selection information of the associated user, sorting by how many times the user has visited;
s5: the user edits and sets the future time period of the current path;
s6: predicting the path influence factors of the editing time period by taking the real-time path information and the historical path data as reference data;
s7: comprehensively sorting according to the cost consumed by the influence factors;
s8: displaying the sequencing result, providing an optimal path and a secondary path, screening out the residual paths, and automatically confirming the optimal path in a default state;
s9: and recording the selection result and generating a work report.
9. The POI recommendation algorithm for learning ranked POI fused with trust relationship of claim 8, wherein the attributes of the path information in step S2 include: current road conditions, current weather conditions, and vehicle restrictions.
10. The POI recommendation algorithm for learning ranked POI fused with trust relationship as claimed in claim 9, wherein the attributes of the work report in step S9 include: recording time, recording location, generated path information, and path selection result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210395703.3A CN114722290A (en) | 2022-04-15 | 2022-04-15 | Trust-relationship-fused ranking learning POI recommendation algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210395703.3A CN114722290A (en) | 2022-04-15 | 2022-04-15 | Trust-relationship-fused ranking learning POI recommendation algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114722290A true CN114722290A (en) | 2022-07-08 |
Family
ID=82244576
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210395703.3A Pending CN114722290A (en) | 2022-04-15 | 2022-04-15 | Trust-relationship-fused ranking learning POI recommendation algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114722290A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117573951A (en) * | 2024-01-16 | 2024-02-20 | 每日互动股份有限公司 | Target user screening method, device, medium and equipment |
-
2022
- 2022-04-15 CN CN202210395703.3A patent/CN114722290A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117573951A (en) * | 2024-01-16 | 2024-02-20 | 每日互动股份有限公司 | Target user screening method, device, medium and equipment |
| CN117573951B (en) * | 2024-01-16 | 2024-04-12 | 每日互动股份有限公司 | Target user screening method, device, medium and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107766462B (en) | Interest point recommendation method based on user preference, social reputation and geographic position | |
| KR100802511B1 (en) | System and method for offering searching service based on topics | |
| WO2019184833A1 (en) | Tourism information recommending method and device | |
| CN106021363B (en) | course recommendation method and device | |
| CN112100529B (en) | Search content ordering method and device, storage medium and electronic equipment | |
| CN109167816A (en) | Information-pushing method, device, equipment and storage medium | |
| CN109582969A (en) | Methodology for Entities Matching, device and electronic equipment | |
| KR102301086B1 (en) | Travel route recommendation system on big data and travel route recommendation method | |
| CN108319603A (en) | Object recommendation method and apparatus | |
| KR20150143971A (en) | Mobile social network service system for processing big data for travel infomation based on location and method for processing of the same | |
| CN111460327B (en) | Method and device for searching for interest, storage medium and computer equipment | |
| CN111141301B (en) | Navigation end point determining method, device, storage medium and computer equipment | |
| CN111191133B (en) | Service search processing method, device and equipment | |
| JP2007219655A (en) | Facility information management system, facility information management method and facility information management program | |
| CN110019645A (en) | Index base construction method, searching method and device | |
| CN113468300A (en) | Intelligent message processing system and method based on WeChat interaction | |
| CN109684548B (en) | Data recommendation method based on user map | |
| CN109492081A (en) | Text information search and information interacting method, device, equipment and storage medium | |
| CN111369294B (en) | Software cost estimation method and device | |
| CN114722290A (en) | Trust-relationship-fused ranking learning POI recommendation algorithm | |
| CN112597389A (en) | Control method and device for realizing article recommendation based on user behavior | |
| CN104615621B (en) | Correlation treatment method and system in search | |
| CN111552787A (en) | Question and answer processing method, device, equipment and storage medium | |
| JP2011501849A (en) | Information map management system and information map management method | |
| CN110377805B (en) | Sensor resource recommendation method based on rapid branch allocation and sorting algorithm |
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 |