CN109284500A - Information transmission system and method based on merchants inviting work process and reading preference - Google Patents
Information transmission system and method based on merchants inviting work process and reading preference Download PDFInfo
- Publication number
- CN109284500A CN109284500A CN201810949454.1A CN201810949454A CN109284500A CN 109284500 A CN109284500 A CN 109284500A CN 201810949454 A CN201810949454 A CN 201810949454A CN 109284500 A CN109284500 A CN 109284500A
- Authority
- CN
- China
- Prior art keywords
- information
- enterprise
- word
- portrait
- carrier
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000008569 process Effects 0.000 title claims abstract description 28
- 230000005540 biological transmission Effects 0.000 title claims description 4
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 38
- 238000013178 mathematical model Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 8
- 238000013461 design Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000013139 quantization Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 5
- 230000007704 transition Effects 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 2
- 238000003032 molecular docking Methods 0.000 claims 1
- 238000012546 transfer Methods 0.000 claims 1
- 238000011002 quantification Methods 0.000 abstract description 5
- 238000005259 measurement Methods 0.000 abstract 1
- 238000007726 management method Methods 0.000 description 19
- 238000005516 engineering process Methods 0.000 description 6
- 238000012544 monitoring process Methods 0.000 description 4
- 230000001737 promoting effect Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 230000005856 abnormality Effects 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 2
- 235000014510 cooky Nutrition 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000035800 maturation Effects 0.000 description 2
- 238000013508 migration Methods 0.000 description 2
- 230000005012 migration Effects 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
-
- 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
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention discloses a kind of information-pushing method based on merchants inviting work process and reading preference, in process of information push, using intelligent elements recognition algorithm and text similarity measurement algorithm, the text informations progress quantification treatment such as information that enterprise's portrait, carrier are drawn a portrait and to be pushed, enterprise's portrait information, carrier information and the information to be pushed are carried out by similarity-rough set using Doc2Vec algorithm, to complete the accurate push about information of inviting outside investment;By the collection and calculating to preference is read, the utilization rate of information is improved.It is characteristic of the invention that push process is intimately associated with merchants inviting work process, pushed information is made precisely to meet the demand of trade and investment promotion scene.
Description
Technical field
The invention belongs to field of information system, especially relate to a kind of letter based on merchants inviting work process and reading preference
Cease supplying system and method.
Background technique
During inviting outside investment, no matter investor or carrier require to obtain the accurate information of a large amount of other side, at present
General method is to obtain correlation by means of the information such as internet search engine, enterprise credibility website, news information inquiry means
Information.During this, traditional information of inviting outside investment inquiry and push mode are had the following problems:
1, information dispersion: due to information dispersion in internet, it is desirable to form complete enterprise's portrait or carrier portrait, need
A large amount of manpowers and time are spent to carry out information collection work.
2, information readability is lower: because on the one hand the push of current information is made using broadcast mode without specific aim in this way
Increased costs are pushed at information, certain harassing and wrecking are on the other hand caused to enterprise.
3, information utilization is lower: investment subject need by continuous screening with match could be in numerous letters of inviting outside investment
The information for meeting itself needs, meeting itself investment intent is found in breath, objectively causes the low of information utilization.
4, information timeliness is poor: the intention of policy and enterprise that government invites outside investment constantly is changing, and practical
Often there is certain hysteresis quality in the information of middle push, leading to the cooperation of government and enterprise, there are some problems.
5. lacking objective reading feedback analysis: after the push of trade and investment promotion broadcast message, push content for enterprise customer whether
It really needs, can not often obtain objective feedback.
6. lacking to enterprise development demand analysis: the premise of the validity of information push is to meet enterprise development demand, is passed
Information of inviting outside investment push mode of uniting lacks the precise positioning to enterprise development demand.
Summary of the invention
In view of this, the present invention propose it is a kind of based on merchants inviting work process and read preference information transmission system and side
Enterprise's portrait, carrier portrait and pushed information are carried out quantification treatment, facilitate and realize machine intelligence understanding and comparison, to enterprise by method
Industry pushes accurate information of inviting outside investment, pushes enterprise investment information to carrier, and push process is intimately associated with merchants inviting work process, makes to push away
The demand for sending accurate information to meet trade and investment promotion scene.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of information-pushing method based on merchants inviting work process and reading preference, comprising:
S1: by mathematical model, enterprise's portrait, carrier portrait and information of inviting outside investment three dedicated participle libraries are established;
S2: enterprise, carrier, information quantization processing;Based on enterprise's big data information, using element extraction algorithm, from difference
Dimension carries out labeling processing to the information that Target Enterprise information, carrier information, needs push;
S3: setting information pushes classification;
S4: information generates;
S5: label filtration and comparison;It is screened and is compared according to pushed information label and enterprise label, carrier tag,
It picks out and is pushed to specific enterprise user and carrier trade and investment promotion personnel with targetedly information.
S6: it calculates user and reads preference, and daily every received information content of user is controlled.
S7: information push.
Further, the method for building up of participle library mathematical model described in step S1 includes:
S101: all words form an observation set, B, M in Chinese, and E, S form a state set, wherein B table
Show that prefix word, E indicate that suffix word, M indicate that word in word, S indicate individual character at word;
S102: by artificial data cleansing and mark work, a large amount of learning samples, as training set are obtained;Training set
In be the multiple samples for having segmented completion for having upper and lower semantic logic;
S103: design function calculates probability, state transition probability in Viterbi algorithm, and observation probability is united simultaneously
The sequence that does well is counted, observation sequence completes building for algorithm model;Wherein observation sequence is sentence to be segmented;State sequence
Column are state corresponding to each word in the sentence, are indicated using the state in the state set;Probability, state turn
Probability is moved, observation probability is the calculated probability of corresponding relationship institute in corpus between state and observation;
Further, step S1 includes: by the method that mathematical model establishes participle library
S201: according to mathematical model, using Viterbi algorithm prediction each newly into each word in sentence under the sentence
Corresponding state;
S202: after algorithm finishes, word can be combined into word, met by design function for state corresponding to each word
State E adds carriage return character, the participle text that can be finally completed.
Further, step S2 needs to convert Chinese word segmentation to mathematical form, i.e. participle vectorization, implementation method packet
It includes:
S301: using Word2Vec algorithm, and the probability that the word occurs is inferred by the N number of word in the front and back of the participle;
S302: parameter is determined using maximum likelihood estimate.
Further, the push of information described in step S3 classification includes that legal pushed information, special appointed information and preference push away
It delivers letters three classifications of breath;
Legal pushed information: the change informations push such as legal item and related important law, policy, detailed rules for the implementation;
Special appointed information: the item and timing node that party especially arranges;
Preference pushed information: meet the information such as company information, carrier information, information on services, the industry information of personal preference
Category information.
Further, information described in step S4 includes:
(1) project proposal for investment both sides in inquiry, link up, bring together, consult, investigate, negotiate, the service of landing, complain and coordinate etc. to recruit
The promise and timing node that each stage of quotient's workflow generates;
(2) according to merchants inviting work process and key node, and cooperation is docked and the process and timing node of project landing,
Different time nodes push different contents;
(3) company information is pushed to trade and investment promotion personnel: according to planning Auto-matching enterprise portrait of promoting trade and investment, being drawn a portrait and examined according to enterprise
Rope company information library pushes search result to trade and investment promotion personnel;
(4) carrier information of inviting outside investment is pushed to enterprise: according to planning Auto-matching enterprise portrait of promoting trade and investment, being drawn a portrait and examined according to enterprise
Rope company information library pushes carrier information of inviting outside investment to the enterprise that enterprise draws a portrait is met;
(5) it is pushed to system user and reads preference information;Preference is read according to user, system pushes same label to user
Information;
(6) publication such as legal item and law, policies and regulations, detailed rules for the implementation, revision and interpretation.
Another aspect of the present invention additionally provides a kind of information push system based on merchants inviting work process and reading preference
System, using four layer framework patterns, including system software layer, running environment layer, architecture, application service;
The application service is equipped with:
Scene obtains: merchants inviting work flow nodes, the variation of legal item, user's reading preference information is obtained, as letter
Breath sends scene condition;
Data statistics: counting user reads data, and clicking rate including same label information reads duration, sharing operation, comments
Valence and comment information;
It sends service: by mobile phone notice, wechat, short message, mail, pushing information required for different scenes to user;
Information bank: the information database for needing to push;
Enterprise library: company information, item information database;
Carrier library: the information database including the various merchant vectors such as area, area, functional areas, garden;
Tag database: the tag database established for enterprise, project, carrier;Including being based on enterprise's big data information, adopt
With element extraction algorithm, the data of labeling are carried out to Target Enterprise information from different dimensions;
Participle library: enterprise's portrait, carrier portrait and the information of inviting outside investment three dedicated participle libraries established by mathematical model.
Further, the application service is additionally provided with:
User service: user can pass through computer, smart phone and plate computer to access this system;
View resource: using the user interface in MVC model;
Service broker: in order to which the user of higher level accesses, allow to access this system by the way of agency service;
Directly access: general user uses this system by the way of directly accessing;
Api interface: the user relatively high for safety requirements such as group users uses this system using API mode.
Further, the architecture includes:
Authentication: authentication is carried out using general user+password login mode;
JDBC:JAVA application service and the direct interface of database and standard;
Inspection script: being to be responsible for data preparation for periodically executing the Shell script of task in (SuSE) Linux OS, grab
Take, remove and application program necessary to other timed tasks;
Generate report: the report for borrowing maturation is shown and Core Generator, and system data display thus generates visualization
Report;
Document management: the concurrent executive capability in order to improve WEB service and application service, load sharing specially set up text
Shelves management server, is responsible for the tasks such as the upload, downloading, migration, duplication of picture, file, GIS, Streaming Media etc.;
Session management: the interaction of the entire session activation of user and computer system is kept to track process;Including right
The management of the interactive information such as session, cookie, token;
Transaction management: the management that data access, service logic are executed;
Log management: including system log, database journal and application log;Hardware, system are soft in record system
Part, application software, the information of database and system problem, while can event to occur in monitoring system;
Rights management: the management including system safety regulation, data access rule, application service access rule etc.;
Abnormality processing: by the modes such as SMS platform, mail, perform script to system, network, application, data exception
Situation is handled.
Compared with the existing technology, present invention has the advantage that
The present invention is by using artificial intelligence semantic understanding algorithm, intelligent elements recognition algorithm and similarity-rough set an etc. system
Enterprise's portrait, carrier portrait and pushed information are carried out quantification treatment, facilitate and realize machine intelligence understanding and comparison by column algorithm,
Accurate information of inviting outside investment is pushed to enterprise, pushes enterprise investment information to carrier.It is characteristic of the invention that push process and trade and investment promotion work
It is intimately associated as process, makes pushed information more personalized, more accurate.
Detailed description of the invention
Fig. 1 is the process of information push logical schematic of the embodiment of the present invention;
Fig. 2 is the technology path schematic diagram of the enterprise information push of the embodiment of the present invention;
Fig. 3 is the system software architecture schematic diagram of the embodiment of the present invention.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
Below in conjunction with attached drawing, the present invention will be described in detail.
One, as shown in Figure 1, detailed process of the invention is described as follows:
1. step 1: establishing enterprise's portrait, carrier portrait and information of inviting outside investment three dedicated participle libraries
Which kind of learnt through actual experiment, if using general participle library, no matter using participle quantization algorithm or similarity operator
Method, final entirety discrimination are not able to satisfy practical demand.Dedicated participle library is only generated, whole discrimination could be improved.
2. step 2: enterprise, carrier, information quantization processing
Based on enterprise's big data information, using element extraction algorithm, label is carried out from different dimensions to Target Enterprise information
Change processing.After being handled by labeling, the abstract concepts such as enterprise's portrait, carrier portrait and information of inviting outside investment are exactly its all feature mark
The set of label.It may be implemented to carry out quantification treatment to the qualitative description of enterprise, carrier and information of inviting outside investment in this way, this is subsequent use
The basis that computer is handled.
Enterprise portrait quantization dimension include:
Serial number | Dimension |
1 | Enterprise's industry class |
2 | Management state and scale |
3 | Enterprise qualification |
4 | Growth requirement |
5 | Investment property class |
6 | Scale of investment |
5 | Investment preference |
Pushed information quantization dimension include.
3. step 3: setting information pushes classification
Including three legal pushed information, special appointed information and preference pushed information classifications.
Legal pushed information: the change informations push such as legal item and related important law, policy, detailed rules for the implementation;
Special appointed information includes: the item and timing node that party especially arranges;
Preference pushed information includes: the company information for meeting personal preference, carrier information, information on services, industry information etc.
Information category information.
4. step 4: information generates
1. including project proposal for investment both sides in inquiry, link up, bring together, consult, investigate, negotiate, the service of landing, complain and coordinate etc.
The promise and timing node that each stage of merchants inviting work process generates.
2. enterprise investment information: according to planning push company information of promoting trade and investment, foundation information inquiry push company information, talking
Project information push.
3. carrier information of inviting outside investment: according to planning push carrier information of promoting trade and investment, foundation query information push carrier information, talking
Project information push.
4. the publication such as law, policies and regulations, detailed rules for the implementation, revision and interpretation.
5. step 5: label filtration and comparison
It is screened and is compared according to pushed information label and enterprise label, carrier tag, picked out with targetedly
Information is pushed to specific enterprise user and carrier trade and investment promotion personnel.
6. step 6: user is calculated using preference
According to user to the reading preference of pushed information, clicking rate, reading duration including same label information, sharing behaviour
Work, evaluation and comment information, and daily every received information content of user is controlled.
7. step 7: information push
By various ways such as mobile phone notice, short message, wechat, mails, information push is realized.
Two, the technology contents that the present invention is embodied include:
1, technology path
The present invention is by using artificial intelligence semantic understanding algorithm, intelligent elements recognition algorithm and similarity-rough set an etc. system
Enterprise's portrait, carrier portrait and pushed information are carried out quantification treatment by column algorithm.Enterprise's portrait, carrier portrait and pushed information
Labeling method extracts semantic prototype using similarity comparison algorithm, labels using elements recognition algorithm to corresponding object
Label.It is the technology path of enterprise's portrait labeling as shown in Figure 2, carrier portrait and the technology path of pushed information are almost the same.
2, the method for library founding mathematical models is segmented
For Chinese word segmentation, there are two types of the forms of word: individual character is at word and multiword at word.For multiword at word, word
Lead-in number-letter relation table B (begin) mark, suffix word can be marked with E (end), and word can be marked with M (middle) in word;For list
Word can be marked at word with S (single).Therefore, each word in Chinese can all correspond to B, M, E, a letter in S,
This is typical Hidden Markov Model.It is considered that word all in Chinese forms an observation set, B, M, E, S composition
One state set, then being Hidden Markov Model third problem -- forecasting problem for the participle in short done.Cause
This solves the problems, such as this using condition random field here.For hidden Markov model third problem, Viterbi algorithm is mesh
The preceding mainstream algorithm for solving the problems, such as this, the exploitation of this Words partition system are also to select the algorithm.If wanting to determine algorithm model, need
Want five parameters, status switch, observation sequence, probability, state transition probability, observation probability.
Observation sequence is sentence to be segmented, and status switch is state corresponding to each word in the sentence, initially
Probability, state transition probability, observation probability are the calculated probability of corresponding relationship institute in corpus between state and observation.
We can obtain a large amount of learning samples, as training set by artificial data cleansing and mark work.In the training set
To there is the multiple samples for having segmented completion of upper and lower semantic logic.For the sample, the design function meter in Viterbi algorithm
Probability, state transition probability, observation probability are calculated, while status switch, observation sequence can be counted, it can completes to calculate
Method model is built.Prediction process is to utilize Viterbi algorithm prediction each newly into each word in sentence under the sentence
Word after algorithm finishes, design function: for state corresponding to each word, can be combined into word, met by corresponding state
State E adds carriage return character, the participle text that can be finally completed.
3, the implementation method of vectorization is segmented
After obtaining participle model, in order to introduce machine learning and deep learning related algorithm, need Chinese word segmentation
It is converted into mathematical form.For the method for transformation used herein for Word2Vec algorithm, algorithm core is mainly the front and back N for passing through the word
A word infers probability that the word occurs, and determines parameter using maximum likelihood estimate.The dimension of term vector in Word2Vec
Degree can be artificially defined.After completing Word2Vec, the term vector of acquisition directly can be and meaningful with doing mathematics operation.
By term vector, the task of all kinds of Chinese natural language processing can be completed.For example, entire chapter text can be calculated using term vector
Or the cosine similarity between sentence, it can differentiate the difference between text or between sentence.For another example, can by word to
Amount builds LSTM (shot and long term Memory Neural Networks), can accomplish text classification.The scientific research institutions such as Google develop after at present
The preferable term vector model of effect, while also thering is relevant framework to build and finishing and appear on the market, therefore by by Chinese word segmentation
Be converted to term vector so as to the method for doing mathematics operation be facts have proved it is feasible.
Three, software system architecture of the invention
As shown in figure 3, this system software uses four layer framework patterns:
1) system software layer
System software layer includes operating system and application server system, and operating system is used based on Linux's at present
CentOS operating system, application server include Apache web server software and MySQL database server software.
2) running environment layer
This system is developed using JAVA language, therefore at runtime, needs JAVA to run basic framework environment, this is
System uses J2EE frame.
3) architecture
Authentication: authentication is carried out using general user+password login mode.
JDBC:JAVA application service and the direct interface of database and standard.
Inspection script: being to be responsible for data preparation for periodically executing the Shell script of task in (SuSE) Linux OS, grab
Take, remove and application program necessary to other timed tasks.
Generate report: the report for borrowing maturation is shown and Core Generator, and system data display thus generates visualization
Report.
Document management: the concurrent executive capability in order to improve WEB service and application service, load sharing specially set up text
Shelves management server, is responsible for the tasks such as the upload, downloading, migration, duplication of picture, file, GIS, Streaming Media etc..
Session management: the interaction of the entire session activation of user and computer system is kept to track process.Including right
The management of the interactive information such as session, cookie, token.
Transaction management: the management that data access, service logic are executed.
Log management: including system log, database journal and application log.Hardware, system are soft in record system
Part, application software, the information of database and system problem, while can event to occur in monitoring system.
Rights management: the management including system safety regulation, data access rule, application service access rule etc..
Abnormality processing: by the modes such as SMS platform, mail, perform script to system, network, application, data exception
Situation is handled.
4) application service
User service: user can pass through computer, smart phone and plate computer to access this system.
View resource: using the user interface in MVC model.
Service broker: in order to which the user of higher level accesses, allow to access this system by the way of agency service.
Directly access: general user uses this system by the way of directly accessing.
Api interface: the user relatively high for safety requirements such as group users can use this system using API mode.
Scene obtains: merchants inviting work flow nodes, the variation of legal item, user's reading preference information is obtained, as letter
Breath sends scene condition;
Data statistics: counting user reads data, and clicking rate including same label information reads duration, sharing operation, comments
Valence and comment information;
It sends service: by mobile phone notice, wechat, short message, mail, pushing information required for different scenes to user;
Operation monitoring: security monitoring is carried out to operation datas such as the access, log, process of system;
Information bank: the information database for needing to push;
Enterprise library: company information, item information database;
Carrier library: the information database including the various merchant vectors such as area, area, functional areas, garden;
Tag database: the tag database established for enterprise, project, carrier;Including being based on enterprise's big data information, adopt
With element extraction algorithm, the data of labeling are carried out to Target Enterprise information from different dimensions;
Participle library: enterprise's portrait, carrier portrait and the information of inviting outside investment three dedicated participle libraries established by mathematical model.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (7)
1. a kind of information-pushing method based on merchants inviting work process and reading preference characterized by comprising
S1: by mathematical model, enterprise's portrait, carrier portrait and information of inviting outside investment three dedicated participle libraries are established;
S2: enterprise, carrier, information quantization processing;Based on enterprise's big data information, using element extraction algorithm, from different dimensions
Labeling processing is carried out to the information that Target Enterprise information, carrier information, needs push;
S3: setting information pushes classification;
S4: information generates;
S5: label filtration and comparison;It is screened and is compared according to pushed information label and enterprise label, carrier tag, selected
It provides targeted information and is pushed to specific enterprise user and trade and investment promotion personnel;
S6: it calculates user and reads preference, and daily every received information content of user is controlled;
S7: information push.
2. the method according to claim 1, wherein segmenting the method for building up packet of library mathematical model described in step S1
It includes:
S101: all words form an observation set, B, M in Chinese, and E, S form a state set, and wherein B indicates word
Lead-in, E indicate that suffix word, M indicate that word in word, S indicate individual character at word;
S102: by artificial data cleansing and mark work, a large amount of learning samples, as training set are obtained;In training set
To there is the multiple samples for having segmented completion of upper and lower semantic logic;
S103: design function calculates probability, state transition probability in Viterbi algorithm, and observation probability counts simultaneously
Status switch, observation sequence complete building for algorithm model;Wherein observation sequence is sentence to be segmented;Status switch is
For state corresponding to each word in the sentence, indicated using the state in the state set;Probability, state transfer are general
Rate, observation probability are the calculated probability of corresponding relationship institute in corpus between state and observation.
3. the method according to claim 1, wherein step S1 establishes the method packet in participle library by mathematical model
It includes:
S201: according to mathematical model, using Viterbi algorithm prediction each, newly into each word in sentence, institute is right under the sentence
The state answered;
S202: after algorithm finishes, word can be combined into word, meet state by design function for state corresponding to each word
E adds carriage return character, the participle text that can be finally completed.
4. the method according to claim 1, wherein step S2 needs are with the Introduction of enterprises in historical data base
Learning sample, vectorization can be segmented with the mathematical form of operation by converting Chinese word segmentation to, implementation method includes:
S301: using Word2Vec algorithm, and the probability that the word occurs is inferred by the N number of word in the front and back of the participle;
S302: parameter is determined using maximum likelihood estimate.
5. the method according to claim 1, wherein the push of information described in step S3 classification includes legal push letter
Three breath, special appointed information and preference pushed information classifications;
Legal pushed information: the change informations push such as legal item and related important law, policy, detailed rules for the implementation;
Special appointed information: the item and timing node that party especially arranges;
Preference pushed information: meet the information classes such as company information, carrier information, information on services, the industry information of personal preference letter
Breath, and the total amount for receiving information to daily each user limits.
6. the method according to claim 1, wherein information described in step S4 includes:
(1) project proposal for investment both sides in inquiry, link up, bring together, consult, investigate, negotiate, the service of landing, complain and the trade and investment promotions work such as coordinate
Make the promise and timing node that each stage of process generates;
(2) different according to merchants inviting work process and key node, and the process and timing node of cooperation docking and project landing
Timing node pushes different contents;
(3) push company information to trade and investment promotion personnel: foundation, which is promoted trade and investment, plans Auto-matching enterprise portrait, is drawn a portrait according to enterprise and retrieves enterprise
Industry information bank pushes search result to trade and investment promotion personnel;
(4) push carrier information of inviting outside investment to enterprise: foundation, which is promoted trade and investment, plans Auto-matching enterprise portrait, is drawn a portrait according to enterprise and retrieves enterprise
Industry information bank pushes carrier information of inviting outside investment to the enterprise that enterprise draws a portrait is met;
(5) it is pushed to specific user and reads preference information;Preference is read according to user, system pushes the letter of same label to user
Breath;
(6) publication such as legal item and law, policies and regulations, detailed rules for the implementation, revision and interpretation.
7. a kind of information transmission system based on merchants inviting work process and reading preference, which is characterized in that use four layer architecture moulds
Formula, including system software layer, running environment layer, architecture layer, application service layer;
The application service layer is equipped with:
Scene obtains: obtaining merchants inviting work flow nodes, the variation of legal item, user's reading preference information, sends out as information
Send scene condition;
Data statistics: counting user reads data, clicking rate including same label information, read duration, sharing operation, evaluation and
Comment information;
It sends service: by mobile phone notice, wechat, short message, mail, pushing information required for different scenes to user;
Information bank: the information database for needing to push;
Enterprise library: company information, item information database;
Carrier library: the information database including the various merchant vectors such as area, area, functional areas, garden;
Tag database: the tag database established for enterprise, project, carrier;Including being based on enterprise's big data information, using wanting
Plain extraction algorithm carries out the data of labeling from different dimensions to Target Enterprise information;
Participle library: enterprise's portrait, carrier portrait and the information of inviting outside investment three dedicated participle libraries established by mathematical model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810949454.1A CN109284500A (en) | 2018-08-20 | 2018-08-20 | Information transmission system and method based on merchants inviting work process and reading preference |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810949454.1A CN109284500A (en) | 2018-08-20 | 2018-08-20 | Information transmission system and method based on merchants inviting work process and reading preference |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109284500A true CN109284500A (en) | 2019-01-29 |
Family
ID=65183288
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810949454.1A Pending CN109284500A (en) | 2018-08-20 | 2018-08-20 | Information transmission system and method based on merchants inviting work process and reading preference |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109284500A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110909247A (en) * | 2019-12-03 | 2020-03-24 | 掌阅科技股份有限公司 | Text information pushing method, electronic equipment and computer storage medium |
CN111210198A (en) * | 2019-12-30 | 2020-05-29 | 广州高企云信息科技有限公司 | Information delivery method and device and server |
CN114741600A (en) * | 2022-04-21 | 2022-07-12 | 深圳市城市产业发展集团有限公司 | Method and device for business recruitment recommendation of enterprises in industrial park |
CN114971664A (en) * | 2021-02-26 | 2022-08-30 | 富联精密电子(天津)有限公司 | Advertisement putting method and related equipment |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040059705A1 (en) * | 2002-09-25 | 2004-03-25 | Wittke Edward R. | System for timely delivery of personalized aggregations of, including currently-generated, knowledge |
CN103164450A (en) * | 2011-12-15 | 2013-06-19 | 腾讯科技(深圳)有限公司 | Method and device for pushing information to target user |
CN104408093A (en) * | 2014-11-14 | 2015-03-11 | 中国科学院计算技术研究所 | News event element extracting method and device |
CN105260858A (en) * | 2015-11-16 | 2016-01-20 | 苏州天地微易智能科技有限公司 | Oa intelligent office system and management method thereof |
CN105989004A (en) * | 2015-01-27 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Information releasing pretreatment method and device |
CN106296299A (en) * | 2016-08-17 | 2017-01-04 | 杭州安通招商技术有限公司 | The storage Factory Building full data management system and method for trade and investment promotion |
CN106682063A (en) * | 2016-10-20 | 2017-05-17 | 北京跃盟科技有限公司 | Advertising information pushing method, device and system |
CN106897285A (en) * | 2015-12-17 | 2017-06-27 | 北京宸瑞国新科技有限公司 | Data Elements extract analysis system and Data Elements extract analysis method |
CN106982256A (en) * | 2017-03-31 | 2017-07-25 | 百度在线网络技术(北京)有限公司 | Information-pushing method, device, equipment and storage medium |
CN108389133A (en) * | 2018-03-19 | 2018-08-10 | 朱将中 | A kind of intelligent auxiliary throws the decision-making technique of Gu |
-
2018
- 2018-08-20 CN CN201810949454.1A patent/CN109284500A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040059705A1 (en) * | 2002-09-25 | 2004-03-25 | Wittke Edward R. | System for timely delivery of personalized aggregations of, including currently-generated, knowledge |
CN103164450A (en) * | 2011-12-15 | 2013-06-19 | 腾讯科技(深圳)有限公司 | Method and device for pushing information to target user |
CN104408093A (en) * | 2014-11-14 | 2015-03-11 | 中国科学院计算技术研究所 | News event element extracting method and device |
CN105989004A (en) * | 2015-01-27 | 2016-10-05 | 阿里巴巴集团控股有限公司 | Information releasing pretreatment method and device |
CN105260858A (en) * | 2015-11-16 | 2016-01-20 | 苏州天地微易智能科技有限公司 | Oa intelligent office system and management method thereof |
CN106897285A (en) * | 2015-12-17 | 2017-06-27 | 北京宸瑞国新科技有限公司 | Data Elements extract analysis system and Data Elements extract analysis method |
CN106296299A (en) * | 2016-08-17 | 2017-01-04 | 杭州安通招商技术有限公司 | The storage Factory Building full data management system and method for trade and investment promotion |
CN106682063A (en) * | 2016-10-20 | 2017-05-17 | 北京跃盟科技有限公司 | Advertising information pushing method, device and system |
CN106982256A (en) * | 2017-03-31 | 2017-07-25 | 百度在线网络技术(北京)有限公司 | Information-pushing method, device, equipment and storage medium |
CN108389133A (en) * | 2018-03-19 | 2018-08-10 | 朱将中 | A kind of intelligent auxiliary throws the decision-making technique of Gu |
Non-Patent Citations (1)
Title |
---|
杨淦: "基于条件随机场模型的中文分词系统研究与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110909247A (en) * | 2019-12-03 | 2020-03-24 | 掌阅科技股份有限公司 | Text information pushing method, electronic equipment and computer storage medium |
CN111210198A (en) * | 2019-12-30 | 2020-05-29 | 广州高企云信息科技有限公司 | Information delivery method and device and server |
CN111210198B (en) * | 2019-12-30 | 2020-08-21 | 广州高企云信息科技有限公司 | Information delivery method and device and server |
CN114971664A (en) * | 2021-02-26 | 2022-08-30 | 富联精密电子(天津)有限公司 | Advertisement putting method and related equipment |
CN114741600A (en) * | 2022-04-21 | 2022-07-12 | 深圳市城市产业发展集团有限公司 | Method and device for business recruitment recommendation of enterprises in industrial park |
CN114741600B (en) * | 2022-04-21 | 2022-12-27 | 深圳市城市产业发展集团有限公司 | Method and device for recommending enterprise business recruitment in industrial park |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20210357378A1 (en) | Multi-service business platform system having entity resolution systems and methods | |
Guzman et al. | A needle in a haystack: What do twitter users say about software? | |
CN104834668B (en) | The position commending system in knowledge based library | |
CN103346957B (en) | A kind of system and method according to contact person's message alteration contact head image expression | |
CN109284500A (en) | Information transmission system and method based on merchants inviting work process and reading preference | |
CN110020660A (en) | Use the integrity assessment of the unstructured process of artificial intelligence (AI) technology | |
CN110598070B (en) | Application type identification method and device, server and storage medium | |
Korzh et al. | Methods for forming an informational image of a higher education institution. | |
CN111885399A (en) | Content distribution method, content distribution device, electronic equipment and storage medium | |
Washizaki et al. | Software-engineering design patterns for machine learning applications | |
CN113902534A (en) | Interactive risk group identification method based on stock community relation map | |
CN116402512B (en) | Account security check management method based on artificial intelligence | |
Dey et al. | Which pull requests get accepted and why? a study of popular npm packages | |
Gerlich et al. | Artificial intelligence as toolset for analysis of public opinion and social interaction in marketing: identification of micro and nano influencers | |
CN111882113B (en) | Enterprise mobile banking user prediction method and device | |
Castillo et al. | Using employee‐generated content from digital platforms to understand the luxury culture | |
CN116595191A (en) | Construction method and device of interactive low-code knowledge graph | |
Impedovo et al. | Next Activity Prediction: An application of shallow learning techniques against deep learning over the BPI Challenge 2020 | |
Wachnik et al. | An analysis of the causes and consequences of the information gap in IT projects. The client’s and the supplier’s perspective in Poland | |
Scholta | Semi-automatic inductive derivation of reference process models that represent best practices in public administrations | |
CN109857753A (en) | User data verification method, device, electronic equipment and storage medium | |
CN107908620A (en) | A kind of method and apparatus based on job documentation anticipation user's occupation | |
CN112148770A (en) | User behavior prediction method and system based on multi-source data and semantic fusion | |
Gupta et al. | An online document emoji-based classification using twitter dataset | |
Doğan et al. | Discovering the same job ads expressed with the different sentences by using hybrid clustering algorithms |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190129 |
|
RJ01 | Rejection of invention patent application after publication |