CN103455596B - A kind of method of science and technology items based on big data project verification assessment - Google Patents
A kind of method of science and technology items based on big data project verification assessment Download PDFInfo
- Publication number
- CN103455596B CN103455596B CN201310393575.XA CN201310393575A CN103455596B CN 103455596 B CN103455596 B CN 103455596B CN 201310393575 A CN201310393575 A CN 201310393575A CN 103455596 B CN103455596 B CN 103455596B
- Authority
- CN
- China
- Prior art keywords
- result
- project
- science
- text
- vector space
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention proposes the method for a kind of science and technology items based on big data project verification assessment, comprises the steps: that acquisition is approved and initiate a project request for data；Compare with project verification request for data after project correlation technique latest development direction and focus are processed and obtain the first result；Compare with project verification request for data after national policy is helped direction and local government's industrial development direction information processing and obtain the second result；The 3rd result is exported after authorities' industry on the spot is investigated feedback suggestion and comprehensive strength information processing；Compare with project verification request for data after similar science and technology item over the years to authorities project verification, established project construction situation over the years and effect process and obtain the 4th result；Project verification application assessment and suggestion report is obtained according to the first result, the second result, the 3rd result, the 4th result and weight thereof.The method implementing science and technology items based on the big data project verification assessment of the present invention, has the advantages that the quality of evaluation of S&T projects and level is higher, assessment result is more objective, more comprehensively.
Description
Technical field
The present invention relates to project evaluation field, particularly to the project verification assessment of a kind of science and technology items based on big data
Method.
Background technology
Science and technology item information has the big data characteristicses such as data type destructuring, data volume be huge, and data are come
Source is often by the trans-regional data accumulated for many years, and the process time using common machines and algorithm is longer.
Existing evaluation of S&T projects uses expert analysis mode pattern, and the main contents of assessment include: evaluation item is set up the project
Necessity and feasibility；Using development present situation, technical situation and the research level of project as index
It is evaluated；Project research and development content, technical matters route, embodiment are evaluated；Estimation items
The ability of mesh carrier, to the capacity of scientific research of carrier, the professional standards of skilled personnel, research equipment
The operating capability of equipment and science and technology item is estimated；The implementation of conditions of evaluation item funds, to funds source,
Self-raised fund, government's supporting funds and dollar amount are evaluated.Owing to science and technology item information has big data
Complexity feature, and the understanding of expert is different, ken has deviation, with certain subjective consciousness,
And application procedures also have complexity, the condition of declaring has ambiguity, so project evaluation is only by specially
The conclusion often scientific basis that family's scoring draws is not enough, and the project that is hardly formed preferably is known together.So expert
Evaluation can exist some random, blindness and one-sidedness, cause quality and the level of evaluation of S&T projects
The highest, science and technology item experts' evaluation mechanism imperfection.Comprehensive and the credibility of expert's project scoring need into
One step is strengthened.
Summary of the invention
The technical problem to be solved in the present invention is, for the quality of the above-mentioned evaluation of S&T projects of prior art
With poor defect, it is provided that the quality of a kind of evaluation of S&T projects and higher based on the big data of level
The method of science and technology item project verification assessment.
The technical solution adopted for the present invention to solve the technical problems is: construct a kind of science and technology based on big data
The method of project evaluation, comprises the steps:
A) science and technology item project verification request for data is obtained；
B) obtain latest development direction and the hot information of science and technology item correlation technique, and it is processed
To current research hot spot data, by the phase of described current research hot spot data with science and technology item project verification request for data
Should partly compare and obtain the first result；
C) obtain national policy and help direction and local government's industrial development direction information, obtain after it is processed
Help bearing data, the appropriate section of described support bearing data with science and technology item project verification request for data is carried out
Relatively obtain the second result；
D) obtain authorities' industry on the spot and investigate the comprehensive strength letter of feedback advisory information and project application unit
Breath, and export the 3rd result embodied with data mode after processing；
E) obtain authorities similar science and technology items over the years project verification situation, established project construction situation the most over the years and
Effect, and obtain a vector space collection to after its process, by described vector space collection and described science and technology item
The appropriate section of project verification request for data compares and obtains the 4th result；
F) described first result, the second result, the 3rd result and the weight of the 4th result are determined, and according to institute
State the first result, the second result, the 3rd result, the 4th result and respective weight and obtain science and technology item project verification
The assessment of application and project verification suggestion report.
In the method for science and technology items based on big data of the present invention project verification assessment, described step A)
Farther include:
A1) the first text about science and technology item project verification application is obtained；
A2) described first text is processed the primary vector space obtained for representing project verification application project；
Described science and technology item project verification request for data is described primary vector space.
In the method for science and technology items based on big data of the present invention project verification assessment, described step B) enter
One step includes:
B1) obtain the latest development direction of science and technology item correlation technique and hot information and form a series of text
Record；
B2) each text entry is carried out successively Chinese word segmentation, filters extraction text feature after stop words；
B3) described text feature is clustered, and extract the primary vector representing current research focus；
B4) described primary vector is compared with described primary vector space, obtain for representing science and technology item
Mesh project verification application and the first result of current techniques focus degree of association.
In the method for science and technology items based on big data of the present invention project verification assessment, described step C) enter
One step includes:
C1) obtain about national policy support direction and the second text of local government's industrial development direction；
C2) described second text is carried out successively Chinese word segmentation, filter stop words after obtain the 3rd text；
C3) by described 3rd text calculating word frequency is obtained current support direction key word；
C4) weight of each key word is distributed equally and builds secondary vector space；
C5) described secondary vector space is compared with described primary vector space, obtain representing described section
Skill is approved and initiate a project and is applied for and the second result of the support direction goodness of fit.
In the method for science and technology items based on big data of the present invention project verification assessment, described step D)
Farther include:
D1) obtain about the comprehensive strength information of project applying unit, request slip in scientific and technological resources investigating system
Position project construction situation over the years and authorities' industry on the spot investigate a series of texts note of feedback advisory information
Record, and form the 4th text；Described comprehensive strength information includes manpower, financial resources, material resources and base information；
D2) from described 4th text, performance information is extracted；
D3) according to described performance information, described applying unit is carried out prestige scoring；
D4) scoring of described prestige is normalized obtains the 3rd result of data type.
In the method for science and technology items based on big data of the present invention project verification assessment, described step E) enter
One step includes:
E1) obtain about authorities' similar science and technology item over the years project verification situation, established project construction over the years
Document data the source C={C1, C2 of situation and effect ... Ci ... }；
E2) from document data source C={C1, C2 ... Ci ... read a text Ci in }；
E3) the word frequency word order prototype vector Vi of described text Ci is initialized；
E4) described text Ci is carried out Chinese word segmentation, and the participle filter that will obtain after described text Ci participle
Except stop words, obtain first participle vector space Ti=(Ti1, Ti2 ..., Tin)；
E5) calculate described first participle vector space Ti=(Ti1, Ti2 ..., Tin) in the word of vector element Tij
Frequently, obtain word frequency weight Fij in corresponding described text Ci, obtain the first word frequency weighing vector space Fi=(Fi1,
Fi2,……,Fin)；
E6) described word frequency weighing vector space Fi is carried out dimensionality reduction, obtain the second word frequency weighing vector space Fi '
=(Fi1, Fi2 ..., Fik) and the second participle vector space Ti '=(Ti1, Ti2 ..., Tik)；
E7) calculate the second participle vector space Ti ' in the word order of vector element, obtain word order weight
Sij (j=1,2 ..., k), and obtain word order weighing vector space S i=(Si1, Si2 ..., Sik)；
E8) the word frequency lexical order vector Vi=(Ti ', Fi ', Si) of text is built；
E9) judge that the text in described document data source C runs through the most, in this way, perform step E10)；
Otherwise, step E2 is returned) read next text；
E10) generate corresponding to described document data source C={C1, C2 ... Ci ... } the word frequency word order of Chinese version
Vector space V={V1, V2 ... Vi}；
E11) by described word frequency lexical order vector SPACE V={ V1, V2 ... Vi} enters with described primary vector space
Row compares, and obtains representing the 4th result of similarity between the project verification application of described science and technology item and project over the years.
In the method for science and technology items based on big data of the present invention project verification assessment, described step F) enter
One step includes:
F1) judging whether the 4th result is more than the threshold value set, in this way, scoring is 0；Otherwise, step is performed
F2)；
F2) described first result, the second result, the 3rd result and the weight of the 4th result are determined；
F3) according to described first result, the second result, the 3rd result, the 4th result and respective weight thereof
Obtain score value；The suggestion report of output project verification simultaneously.
In the method for science and technology items based on big data of the present invention project verification assessment, described project verification suggestion
Report includes similarity the setting up the project over the years in set point with science and technology item project verification application, declares unit phase
Like project, unit performance situation, current support direction and key technology hot information.
In the method for science and technology items based on big data of the present invention project verification assessment, described step A2)
In primary vector space be according to described E2) to E11) and method obtain.
The method implementing science and technology items based on the big data project verification assessment of the present invention, has the advantages that
Due to by standing with science and technology item after the latest development direction of science and technology item correlation technique and hot information are processed
The appropriate section of request for data compares and obtains the first result；By national policy being helped direction and local political affairs
After the industrial development direction information processing of mansion, the appropriate section with science and technology item project verification request for data compares and obtains second
Result；By authorities' industry on the spot is investigated feedback advisory information and the comprehensive strength of project application unit
The 3rd result is exported after information processing；By similar science and technology item over the years to authorities project verification situation, over the years
With the appropriate section ratio of described science and technology item project verification request for data after established project construction situation and effect process
Relatively obtain the 4th result；By the first result, the second result, the 3rd result, the 4th result and respective weight
Obtain assessment and the project verification suggestion report of science and technology item project verification application, so being made that approving and initiate a project and comparing
More scientific and normal careful assessment, experts' evaluation be referred to use for reference model evaluation with further comprehensively, carefully,
Understand by the concrete condition of Project evaluation, it is to avoid some in experts' evaluation are random, blind in the past responsiblely
From property and the drawback of one-sidedness, so that the quality of evaluation of S&T projects and level are higher.
Accompanying drawing explanation
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to enforcement
In example or description of the prior art, the required accompanying drawing used is briefly described, it should be apparent that, describe below
In accompanying drawing be only some embodiments of the present invention, for those of ordinary skill in the art, do not paying
On the premise of going out creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the flow process in one embodiment of method of present invention science and technology item based on big data project verification assessment
Figure；
Fig. 2 is the particular flow sheet obtaining science and technology item project verification request for data in described embodiment；
Fig. 3 is that in described embodiment, latest development direction and hot information to science and technology item correlation technique is carried out
The particular flow sheet processed；
Fig. 4 is to help direction and local government's industrial development direction information to enter national policy in described embodiment
The particular flow sheet that row processes；
Fig. 5 is, in described embodiment, authorities' industry on the spot is investigated feedback advisory information and Project Unit
Comprehensive strength information carries out the particular flow sheet processed；
Fig. 6 is the concrete stream that in described embodiment, similar science and technology item situation over the years to authorities processes
Cheng Tu；
Fig. 7 is the assessment of science and technology item project verification application in described embodiment and the concrete of suggestion report acquisition of setting up the project
Flow chart.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clearly
Chu, be fully described by, it is clear that described embodiment be only a part of embodiment of the present invention rather than
Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creation
The every other embodiment obtained under property work premise, broadly falls into the scope of protection of the invention.
In the embodiment of the method for present invention science and technology item based on big data project verification assessment, the flow process of its method
Figure is as shown in Figure 1.In Fig. 1, the method comprises the steps:
Step S01 obtains science and technology item project verification request for data: in this step, obtains science and technology item project verification application
Data, namely obtain the data about new established project.About the most specifically obtaining science and technology item project verification application
Data, will be described in detail after a while.
Step S02 obtains latest development direction and the hot information of science and technology item correlation technique, and carries out it
Process obtains current research hot spot data, by current research hot spot data and science and technology item project verification request for data
Appropriate section compares and obtains the first result: in this step, for each Large-scale professional technology web sites, periodical
Carry out statistical analysis with the relevant big data in scientific and technological resources investigation and management system, use compound recipe Chinese word segmentation to calculate
Relevant big data are processed by method and Kmeans clustering algorithm, export the forward position research heat of current scientific research technology
Point data, compares the appropriate section of current research hot spot data with science and technology item project verification request for data
To the first result.Thus achieve the latest development direction to science and technology item correlation technique and hot information
Objective evaluation.
Step S03 obtains national policy and helps direction and local government's industrial development direction information, processes it
After obtain helping bearing data, enter helping the set up the project appropriate section of request for data of bearing data and science and technology item
Go to compare and obtain the second result: in this step, obtain national policy and help direction and local government's industry development
Directional information, its main information is that national policy is helped, the industry development of local government is sent to and helps direction
Official document, belongs to Text Information Data and excavates, and obtains helping bearing data, will help direction number after processing it
Comparing according to the appropriate section with science and technology item project verification request for data and obtain the second result, the second result is for holding up
Hold direction key word.Thus achieve the national policy to project and help direction, local government's industry development
Help the objective evaluation in direction.
Step S04 obtains authorities industries on the spot and investigates the comprehensive of feedback advisory information and project application unit
Information for strength, and export the 3rd result embodied with data mode after processing: in this step, obtain main
Pipe department industry on the spot investigates feedback advisory information and the comprehensive strength information of project application unit, and it is mainly believed
Breath is authorities' on-the-spot investigation local industries development, investigates engineering center at different levels and key lab of enterprise
Feedback advisory information, exports the 3rd result embodied with data mode after processing it, the 3rd result is for helping
Direction keyword and the support dynamics factor.Thus achieve and authorities' industry on the spot is investigated feedback opinion
Objective evaluation.
Step S05 obtains authorities' similar science and technology item over the years project verification situation, established project construction over the years
Situation and effect, and obtain a vector space collection to after its process, vector space collection is stood with science and technology item
The appropriate section of item request for data compares and obtains the 4th result: in this step, obtains science and technology integrated service
Authorities in management system established project situation, approving and initiate a project over the years of project application unit over the years is built
If situation and performance information, use " high-performance calculation text based on TF/IDF and Markov model spy
Levy extraction algorithm " (The high performance computing text feature extraction
Algorithm based on the TF/IDF and markov model) (it is called for short:
HpTF/IDF-MM) these information are carried out process and obtain a vector space collection, by vector space collection and section
The approve and initiate a project appropriate section of request for data of skill compares and obtains the 4th result.Thus achieve supervisor
Department's similar terms over the years project verification situation, established project construction situation and the objective evaluation of performance information over the years.
Step S06 determines the first result, the second result, the 3rd result and the weight of the 4th result, and foundation
First result, the second result, the 3rd result, the 4th result and respective weight obtain science and technology item project verification Shen
Assessment please and project verification suggestion report: in this step, determine the first result, the second result, the 3rd result and
The weight of the 4th result, in the present embodiment, the setting of weight is configured by rule of thumb by expert, then foundation
First result, the second result, the 3rd result, the 4th result and respective weight obtain science and technology item project verification Shen
Assessment please and project verification suggestion report.Describe for convenience, in the present embodiment, by the first result, the second knot
Really, the 3rd result and the 4th result are respectively labeled as M1, M2, M3 and M4, then stand according to project over the years
Expert's assessment determine the weight of M1, M2, M3 and M4, the final assessment obtaining project verification application with
And project verification suggestion report.
Project verification suggestion report provides relevant project verification to advise reference, project verification suggestion report with the pattern of text to expert
The content accused includes: with the similarity of science and technology item project verification application project verification (this reality over the years in set point
Execute in example as the project verification over the years of 5 before similarity), declare unit similar terms, unit performance situation, currently hold up
Hold direction and key technology hot information.Wherein, 5 set up the project as reference information over the years before similarity；Defeated
Go out to declare unit similar terms for avoiding repeating project verification.
In the present embodiment, the project obtained by step S01-S06 is marked, and marks the highest, feasibility of setting up the project
The best；Current support direction and focus key technology is understood by step S02, S03；Carried by step S04
For the credit rating of unit, comprehensive strength and performance situation；It is extracted first 5 of similarity by step S05
Project name, Distribution Area.Science and technology item expert analysis mode pattern in compared to existing technology, the present invention couple
Appraisement system that science and technology item project verification is estimated having quantization and relatively uniform evaluation criterion, by section
The recent technological advances direction of skill project, hot technology；National policy helps direction；Local government's industry is sent out
Direction is helped in exhibition；Authorities' similar terms over the years project verification situation；Established project construction situation the most over the years with become
Effect and authorities' industry on the spot investigate the scoring of six aspects of feedback opinion, make approving and initiate a project relatively
Relatively scientific and normal careful assessment.Experts' evaluation be referred to use for reference appraisal procedure to accomplish further comprehensively,
Carefully, understand by the concrete condition of Project evaluation, it is to avoid some in current experts' evaluation are random responsiblely
Property, following blindly property and one-sidedness such that it is able to play the effect improving science and technology item experts' evaluation mechanism.
It is noted that in FIG, for the sake of narration, step S02-S05 has certain
Sequentially.It practice, above-mentioned steps can also be divided into several groups, each group of internal step has certain
Sequentially, between group and group, can be ordering relation as above, it is also possible to be parallel, also may be used
To be the order different from said sequence.Such as, a kind of method of packet is: step S02 and S03 are one
Group, step S04 and S05 are one group, and between these groups can be arranged side by side or executed in parallel, it is possible to
To perform (can not be above-mentioned order) according to certain rule order.
For the present embodiment, above-mentioned steps S01 also can refine further, the idiographic flow after its refinement
Figure is as shown in Figure 2.In Fig. 2, step S01 farther includes:
Step S11 obtains the first text about science and technology item project verification application: in this step, obtains about section
Skill approve and initiate a project application the first text, namely obtain about new projects project verification application the first text.
It is empty that step S12 processes the primary vector obtained for representing project verification application project to the first text
Between: in this step, in this step, the first text is processed and obtains for representing project verification application project
Primary vector space, particularly as be the project verification application to new projects use HpTF/IDF-MM algorithm represented
The primary vector space of new projects, in the present embodiment, is designated as L by primary vector space；Above-mentioned science and technology item stands
Item request for data is primary vector space.HpTF/IDF-MM algorithm can be introduced after a while.
For the present embodiment, above-mentioned steps S02 also can refine further, the idiographic flow after its refinement
Figure is as shown in Figure 3.In Fig. 3, step S02 farther includes:
Step S21 obtains the latest development direction of science and technology item correlation technique and hot information and is formed a series of
Text entry: in this step, the latest development direction of science and technology item correlation technique and hot information are from data
Source is extracted.Data source is country and the science and technology plan item data of provinces and cities, then extracts from data source
Just obtain science and technology item be correlated with about the recent technological advances direction of science and technology item, the data set of hot technology
The latest development direction of technology and hot information, and form a series of text entry.
Step S22 carries out Chinese word segmentation successively, filters extraction text feature after stop words each text entry:
In this step, each text entry is carried out Chinese word segmentation, filters stop words, then extract text feature.
Text feature is clustered by step S23, and extracts the primary vector representing current research focus: this
In step, use Kmeans clustering algorithm to cluster to text feature, and extract expression current research heat
The primary vector of point.
Primary vector is compared by step S24 with primary vector space, obtains for representing that science and technology item stands
Item application and the first result of current techniques focus degree of association: in this step, by primary vector and primary vector
Space L compares, and obtains for representing science and technology item project verification application and the of current techniques focus degree of association
One result M1；The span of the first result M1 is: 0 ＜ M1 ＜ 1.
For the present embodiment, above-mentioned steps S03 also can refine further, the idiographic flow after its refinement
Figure is as shown in Figure 4.In Fig. 4, step S03 farther includes:
Step S31 obtains about national policy support direction and the second text of local government's industrial development direction:
In this step, national policy helps the second text of direction and local government's industrial development direction to be from data source
Middle extraction.Data source is country and the science and technology plan item data of provinces and cities, then extracts from data source and closes
The data set helping direction and local government's industrial development direction in national policy has just obtained national policy support
Direction and the second text of local government's industrial development direction.
Step S32 carries out Chinese word segmentation successively to the second text, filter stop words after obtain the 3rd text: this
In step, the first text is carried out successively Chinese word segmentation, filter the pretreatment such as stop words after obtain the 3rd text.
Step S33 obtains current support direction key word by the 3rd text calculates word frequency: in this step,
Current support direction key word is obtained by the 3rd text being calculated word frequency.
Step S34 is distributed equally and builds secondary vector space to the weight of each key word: in this step,
The weight of each key word is distributed equally, and direct construction vector space model, in the present embodiment, will
This vector space model is referred to as secondary vector space.
Secondary vector space is compared by step S35 with primary vector space, obtains representing that science and technology item stands
Item application is with the second result helping the direction goodness of fit: in this step, by secondary vector space and primary vector
Space L compares, and obtains the second result M2 representing science and technology item project verification application with helping the direction goodness of fit,
The span of the second result M2 is: 0 ＜ M2 ＜ 1.
For the present embodiment, above-mentioned steps S04 also can refine further, the idiographic flow after its refinement
Figure is as shown in Figure 5.In Fig. 5, step S04 farther includes:
Step S41 obtains about the comprehensive strength information of project applying unit, Shen in scientific and technological resources investigating system
Please unit project construction over the years situation and a series of texts of authorities' investigation feedback advisory information of industry on the spot
Record, and form the 4th text: in this step, extract from above-mentioned data source and investigate system about scientific and technological resources
The comprehensive strength information of Tong Zhong project applying unit, applying unit's project construction over the years situation and authorities are real
Real estate sector investigates a series of text entries of feedback advisory information, and forms the 4th text, it is worth mentioning at this point that,
In the present embodiment, comprehensive strength information includes manpower, financial resources, material resources and base information.
Step S42 extracts performance information from the 4th text: in this step, extracts performance from the 4th text
Information.
Step S43 carries out prestige scoring to applying unit according to performance information: in this step, by identifying the
Performance information in four texts, can determine the situation of passing judgement on, then applying unit be carried out prestige scoring.
Step S44 is normalized the 3rd result obtaining data type to prestige scoring: in this step,
Prestige scoring is normalized, is output as the 3rd result M3 of data type.Taking of 3rd result M3
Value scope is: 0 ＜ M3 ＜ 1.
For the present embodiment, in above-mentioned steps S05, the situation that main input is approved and initiate a project over the years, its
Including the autograph of project, unit, application documents, report of feasibility etc., data source is arranged, will be every
Individual item is visually every text entry.Text entry is carried out Chinese word segmentation, filters stop words, structure text
Feature, set up vector space, use high-performance calculation text feature based on TF/IDF and Markov model
Extraction algorithm (HpTF/IDF-MM) builds vector space collection based on word frequency word order, and its output result is one
The vector space collection of system.Contrast with primary vector space L, take the value of highest similarity, be designated as the 4th knot
Really M4.The particular flow sheet of step S05 is as shown in Figure 6.In Fig. 6, step S05 farther includes:
Step S501 obtains about authorities' similar science and technology item over the years project verification situation, established project over the years
Document data the source C={C1, C2 of construction situation and effect ... Ci ... }: in this step, obtain about supervisor
Department's similar science and technology item over the years project verification situation, established project construction situation and the document data of effect over the years
Source C={C1, C2 ... Ci ... }.
Step S502 is from document data source C={C1, C2 ... Ci ... read text Ci: this step in }
In, from document data source C={C1, C2 ... Ci ... read a text Ci in }.
Step S503 initializes the word frequency word order prototype vector Vi of text Ci: in this step, initializes text
The word frequency word order prototype vector Vi of Ci.
Step S504 carries out Chinese word segmentation to text Ci, and is filtered by the participle obtained after text Ci participle and stop
Word, obtain first participle vector space Ti=(Ti1, Ti2 ..., Tin): in this step, text Ci is carried out
Chinese word segmentation, and the participle obtained after text Ci participle is filtered stop words, obtain first participle vector space
Ti=(Ti1, Ti2 ..., Tin).
Step S505 calculate first participle vector space Ti=(Ti1, Ti2 ..., Tin) in the word of vector element Tij
Frequently, obtain word frequency weight Fij in corresponding text Ci, obtain the first word frequency weighing vector space Fi=(Fi1,
Fi2 ..., Fin): in this step, calculate first participle vector space Ti=(Ti1, Ti2 ..., Tin) in element vector
The word frequency of element Tij, obtains word frequency weight Fij in corresponding text Ci, obtains the first word frequency weighing vector space
Fi=(Fi1, Fi2 ..., Fin).
Step S506 carries out dimensionality reduction to word frequency weighing vector space Fi, obtains the second word frequency weighing vector space
Fi '=(Fi1, Fi2 ..., Fik) and the second participle vector space Ti '=(Ti1, Ti2 ..., Tik): this step
In Zhou, word frequency weighing vector space Fi is carried out dimensionality reduction, obtain the second word frequency weighing vector space Fi '=(Fi1,
Fi2 ..., Fik) and the second participle vector space Ti '=(Ti1, Ti2 ..., Tik).
Step S507 calculates the second participle vector space Ti ' in the word order of vector element, obtain word order weight
Sij (j=1,2 ..., k), and obtain word order weighing vector space S i=(Si1, Si2 ..., Sik): in this step,
Calculate the second participle vector space Ti ' in the word order of vector element, obtain word order weight Sij (j=1,2 ..., k),
And obtain word order weighing vector space S i=(Si1, Si2 ..., Sik).
Step S508 builds the word frequency lexical order vector Vi=(Ti ', Fi ', Si) of text: in this step, builds literary composition
Word frequency lexical order vector Vi=originally (Ti ', Fi ', Si).
Step S509 judges that the text in the C of document data source runs through the most: in this step, it is judged that number of files
The most run through according to the text in the C of source, in this way, perform step S510；Otherwise, return step S502 to read
Take next text.
Step S510 generates corresponding to document data source C={C1, C2 ... Ci ... } the word frequency word order of Chinese version
Vector space V={V1, V2 ... Vi}: if the judged result of above-mentioned steps S509 is yes, then perform basis
Step.In this step, generate corresponding to document data source C={C1, C2 ... Ci ... } the word frequency word of Chinese version
Ordered vector space V={V1, V2 ... Vi}.
Step S511 is by word frequency lexical order vector SPACE V={ V1, V2 ... Vi} compares with primary vector space
Relatively, obtain representing the 4th result of similarity between science and technology item project verification application and project over the years: in this step,
By word frequency lexical order vector SPACE V={ V1, V2 ... Vi} compares with primary vector space L, is represented
4th result M1 of similarity between science and technology item project verification application and project over the years.The value of the 4th result M4
Scope is: 0 ＜ M4 ＜ 1.
It is noted that the primary vector space L in above-mentioned steps S12 is according to above-mentioned steps S512 extremely
The method of step S511 obtains.
HpTF/IDF-MM algorithm is applicable to High-Performance Computing Cluster application, uses Mapreduce programming, it is possible to
The big Data processing of project material reaches effect rapidly and efficiently.Input information is text document data collection, defeated
Go out the document space vector set after information is characterized weighting.Algorithm can extract good characteristic information, builds
Representative vector space, it is particular in that it can either extract text by word frequency well
Key character, the feature of every terms of information in project verification material can be reflected by word order again.HpTF/IDF-MM
Algorithm carries out feature extraction to text, can balance the load well in the Text character extraction stage, improves computing
Speed, shortens operation time；This algorithm fusion feature of word frequency and word order, can reflect the concern of project
Focus, the most also can be highly suitable for by the related information between word order reflection science and technology item declaration material
Science and technology item material destructuring but the most normalized data characteristics；This algorithm is realizing in step, first
First with the feature that word frequency amount of calculation is less, the result after HpTF/IDF computing is carried out text dimensionality reduction, then
Carry out the computing of word order feature again, effectively reduce the time complexity of algorithm.
For the present embodiment, above-mentioned steps S06 also can refine further, the idiographic flow after its refinement
Figure is as shown in Figure 7.In Fig. 7, step S06 farther includes:
Step S61 judges that whether the 4th result is more than the threshold value set: in this step, it is judged that the 4th result
Whether M4 is more than the threshold value set, and in this way, performs step S62；Otherwise, step S63 is performed.
The scoring of step S62 is 0: if the judged result of above-mentioned steps S61 is yes, then perform this step.This
In embodiment, scoring (project scoring) is labeled as in F, this step, when M4 is more than the threshold value set
(threshold value of this setting is derived as 0.9 by expertise in training set, it is also possible to freely arrange), project is marked
F=0.
Step S63 determines the first result, the second result, the 3rd result and the weight of the 4th result: on if
The judged result stating step S61 is no, then perform this step.In this step, by the first result, the second knot
Really, the weight of the 3rd result and the 4th result be respectively labeled as W1, W2, W3 and W4, weight is also arranged
It is configured by expertise.Under normal circumstances, the meaning represented according to all types of data, each weight should expire
Following condition: the W1=W2 < W3 < W4 of foot, meanwhile, W1+W2 < 0.5.
Step S64 is according to the first result, the second result, the 3rd result, the 4th result and respective weight thereof
Obtain score value；The suggestion report of output project verification simultaneously: in this step, according to the first result M1, the second result
M2, the 3rd result M3, the 4th result M4 and respective weight thereof obtain score value.Project is marked
F=W1*M1+W2*M2+W3*M3+W4* (1-M4).In this step, the most also output project verification suggestion report
Its content includes: 5 set up the project as reference information over the years before similarity；Unit similar terms is declared in output,
To avoid repeating project verification；Output unit performance situation；Output is current helps direction and focus key technology.
In a word, in the present embodiment, the method is to the recent technological advances direction of science and technology item, hot technology；
National policy helps direction；Direction is helped in local government's industry development；Authorities' similar terms over the years project verification
Situation；Established project construction situation and effect and authorities industry on the spot over the years investigates feedback opinion six
Aspect is estimated, so that " high-performance calculation Text character extraction based on TF/IDF and Markov model is calculated
Method " big data processing technique excavates the factor of marking, objectively to science and technology item from magnanimity science and technology item information
Mark, be aided with expert analysis mode pattern, make evaluation of S&T projects accomplish science, reasonable, symbol as far as possible
Close practical situation, and blindness and the randomness of Authorize to Invest can be reduced for the service of relevant authorities,
There is emphasis, distribute scientific and technological resources efficiently, improve, improve quality and the level of science and technology decision-making.Compare section
Skill project expert analysis mode pattern, the present invention be estimated having to science and technology item project verification quantization appraisement system and
Relatively uniform evaluation criterion, makes the most scientific and normal careful assessment to approving and initiate a project.
The foregoing is only presently preferred embodiments of the present invention, not in order to limit the present invention, all at this
Within bright spirit and principle, any modification, equivalent substitution and improvement etc. made, should be included in this
Within bright protection domain.
Claims (4)
1. the method for science and technology items based on a big data project verification assessment, it is characterised in that include walking as follows
Rapid:
A) science and technology item project verification request for data is obtained；
B) obtain latest development direction and the hot information of science and technology item correlation technique, and it is processed
To current research hot spot data, by the phase of described current research hot spot data with science and technology item project verification request for data
Should partly compare and obtain the first result；
C) obtain national policy and help direction and local government's industrial development direction information, obtain after it is processed
Help bearing data, the appropriate section of described support bearing data with science and technology item project verification request for data is carried out
Relatively obtain the second result；
D) obtain authorities' industry on the spot and investigate the comprehensive strength letter of feedback advisory information and project application unit
Breath, and export the 3rd result embodied with data mode after processing；
E) obtain authorities similar science and technology items over the years project verification situation, established project construction situation the most over the years and
Effect, and obtain a vector space collection to after its process, by described vector space collection and described science and technology item
The appropriate section of project verification request for data compares and obtains the 4th result；
F) described first result, the second result, the 3rd result and the weight of the 4th result are determined, and according to institute
State the first result, the second result, the 3rd result, the 4th result and respective weight and obtain science and technology item project verification
The assessment of application and project verification suggestion report；
Described step A) farther include:
A1) the first text about science and technology item project verification application is obtained；
A2) described first text is processed the primary vector space obtained for representing project verification application project；
Described science and technology item project verification request for data is described primary vector space；
Described step B) farther include:
B1) obtain the latest development direction of science and technology item correlation technique and hot information and form a series of text
Record；
B2) each text entry is carried out successively Chinese word segmentation, filters extraction text feature after stop words；
B3) described text feature is clustered, and extract the primary vector representing current research focus；
B4) described primary vector is compared with described primary vector space, obtain for representing science and technology item
Mesh project verification application and the first result of current techniques focus degree of association；
Described step C) farther include:
C1) obtain about national policy support direction and the second text of local government's industrial development direction；
C2) described second text is carried out successively Chinese word segmentation, filter stop words after obtain the 3rd text；
C3) by described 3rd text calculating word frequency is obtained current support direction key word；
C4) weight of each key word is distributed equally and builds secondary vector space；
C5) described secondary vector space is compared with described primary vector space, obtain representing described section
Skill is approved and initiate a project and is applied for and the second result of the support direction goodness of fit；
Described step D) farther include:
D1) obtain about the comprehensive strength information of project applying unit, request slip in scientific and technological resources investigating system
Position project construction situation over the years and authorities' industry on the spot investigate a series of texts note of feedback advisory information
Record, and form the 4th text；Described comprehensive strength information includes manpower, financial resources, material resources and base information；
D2) from described 4th text, performance information is extracted；
D3) according to described performance information, described applying unit is carried out prestige scoring；
D4) scoring of described prestige is normalized obtains the 3rd result of data type；
Described step E) farther include:
E1) obtain about authorities' similar science and technology item over the years project verification situation, established project construction over the years
Document data the source C={C1, C2 of situation and effect ... Ci ... }；
E2) from document data source C={C1, C2 ... Ci ... read a text Ci in }；
E3) the word frequency word order prototype vector Vi of described text Ci is initialized；
E4) described text Ci is carried out Chinese word segmentation, and the participle filter that will obtain after described text Ci participle
Except stop words, obtain first participle vector space Ti=(Ti1, Ti2 ..., Tin)；
E5) calculate described first participle vector space Ti=(Ti1, Ti2 ..., Tin) in the word of vector element Tij
Frequently, obtain word frequency weight Fij in corresponding described text Ci, obtain the first word frequency weighing vector space Fi=(Fi1,
Fi2,……,Fin)；
E6) described word frequency weighing vector space Fi is carried out dimensionality reduction, obtain the second word frequency weighing vector space Fi '
=(Fi1, Fi2 ..., Fik) and the second participle vector space Ti '=(Ti1, Ti2 ..., Tik)；
E7) calculate the second participle vector space Ti ' in the word order of vector element, obtain word order weight
Sij (j=1,2 ..., k), and obtain word order weighing vector space S i=(Si1, Si2 ..., Sik)；
E8) the word frequency lexical order vector Vi=(Ti ', Fi ', Si) of text is built；
E9) judge that the text in described document data source C runs through the most, in this way, perform step E10)；
Otherwise, step E2 is returned) read next text；
E10) generate corresponding to described document data source C={C1, C2 ... Ci ... } the word frequency word order of Chinese version
Vector space V={V1, V2 ... Vi}；
E11) by described word frequency lexical order vector SPACE V={ V1, V2 ... Vi} enters with described primary vector space
Row compares, and obtains representing the 4th result of similarity between the project verification application of described science and technology item and project over the years.
The method of science and technology items based on big data the most according to claim 1 project verification assessment, its feature
It is, described step F) farther include:
F1) judging whether the 4th result is more than the threshold value set, in this way, scoring is 0；Otherwise, step is performed
F2)；
F2) described first result, the second result, the 3rd result and the weight of the 4th result are determined；
F3) according to described first result, the second result, the 3rd result, the 4th result and respective weight thereof
Obtain score value；The suggestion report of output project verification simultaneously.
The method of science and technology items based on big data the most according to claim 2 project verification assessment, its feature
Being, described project verification suggestion report includes and similarity the going through in set point of science and technology item project verification application
Set up the project in year, declare unit similar terms, unit performance situation, current support direction and key technology focus letter
Breath.
The method of science and technology items based on big data the most according to claim 3 project verification assessment, its feature
Be, described step A2) in primary vector space be according to described E2) to E11) and method obtain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310393575.XA CN103455596B (en) | 2013-09-02 | 2013-09-02 | A kind of method of science and technology items based on big data project verification assessment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310393575.XA CN103455596B (en) | 2013-09-02 | 2013-09-02 | A kind of method of science and technology items based on big data project verification assessment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103455596A CN103455596A (en) | 2013-12-18 |
| CN103455596B true CN103455596B (en) | 2016-11-02 |
Family
ID=49737959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310393575.XA Active CN103455596B (en) | 2013-09-02 | 2013-09-02 | A kind of method of science and technology items based on big data project verification assessment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103455596B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104133842A (en) * | 2014-06-24 | 2014-11-05 | 国家电网公司 | Data processing method and data processing system with intelligent expert detection function |
| CN104156386A (en) * | 2014-06-24 | 2014-11-19 | 国家电网公司 | Data processing method and system with image recognition function |
| CN104133838A (en) * | 2014-06-24 | 2014-11-05 | 国家电网公司 | Data processing method and system with system detection function |
| CN104077485B (en) * | 2014-06-30 | 2017-05-03 | 西安电子科技大学 | Model correctness evaluation method based on goodness of fit |
| CN109636352A (en) * | 2018-12-20 | 2019-04-16 | 湖南晖龙集团股份有限公司 | A kind of distributed content duplicate checking early warning system based on financial big data |
| CN110046225B (en) * | 2019-04-16 | 2020-11-24 | 广东省科技基础条件平台中心 | Scientific and technological project material integrity assessment decision model training method |
| CN110310048B (en) * | 2019-07-10 | 2023-07-04 | 云南电网有限责任公司电力科学研究院 | Distribution network planning overall process evaluation method and device |
| CN111160778A (en) * | 2019-12-30 | 2020-05-15 | 佰聆数据股份有限公司 | Outbound project auditing and evaluating method and system based on big data and computer equipment |
| CN111598331B (en) * | 2020-05-13 | 2023-07-07 | 中国科学院计算机网络信息中心 | Project feasibility prediction analysis method based on scientific research multidimensional features |
| CN113159537B (en) * | 2021-04-06 | 2023-05-23 | 南方电网能源发展研究院有限责任公司 | Assessment method and device for new technical project of power grid and computer equipment |
| CN113421026A (en) * | 2021-07-19 | 2021-09-21 | 首都医科大学附属北京儿童医院 | Hospital scientific research project application management method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7359865B1 (en) * | 2001-11-05 | 2008-04-15 | I2 Technologies Us, Inc. | Generating a risk assessment regarding a software implementation project |
| US7366680B1 (en) * | 2002-03-07 | 2008-04-29 | Perot Systems Corporation | Project management system and method for assessing relationships between current and historical projects |
| CN101697217A (en) * | 2009-11-06 | 2010-04-21 | 金蝶软件（中国）有限公司 | Method and device for generating evaluation scheme |
| CN102402732A (en) * | 2010-09-14 | 2012-04-04 | 中国船舶工业综合技术经济研究院 | Method and system for evaluating scientific research projects |
| CN102426603A (en) * | 2011-11-11 | 2012-04-25 | 任子行网络技术股份有限公司 | Text information regional recognition method and device |
| CN103226751A (en) * | 2013-03-29 | 2013-07-31 | 广西电网公司电力科学研究院 | Post-evaluation method for science and technology projects of grid enterprise |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8019793B2 (en) * | 2003-02-14 | 2011-09-13 | Accenture Global Services Limited | Methodology infrastructure and delivery vehicle |
| US8280897B2 (en) * | 2009-01-30 | 2012-10-02 | Accenture Global Services Limited | Methods and systems for assessing project management offices |
-
2013
- 2013-09-02 CN CN201310393575.XA patent/CN103455596B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7359865B1 (en) * | 2001-11-05 | 2008-04-15 | I2 Technologies Us, Inc. | Generating a risk assessment regarding a software implementation project |
| US7366680B1 (en) * | 2002-03-07 | 2008-04-29 | Perot Systems Corporation | Project management system and method for assessing relationships between current and historical projects |
| CN101697217A (en) * | 2009-11-06 | 2010-04-21 | 金蝶软件（中国）有限公司 | Method and device for generating evaluation scheme |
| CN102402732A (en) * | 2010-09-14 | 2012-04-04 | 中国船舶工业综合技术经济研究院 | Method and system for evaluating scientific research projects |
| CN102426603A (en) * | 2011-11-11 | 2012-04-25 | 任子行网络技术股份有限公司 | Text information regional recognition method and device |
| CN103226751A (en) * | 2013-03-29 | 2013-07-31 | 广西电网公司电力科学研究院 | Post-evaluation method for science and technology projects of grid enterprise |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103455596A (en) | 2013-12-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103455596B (en) | A kind of method of science and technology items based on big data project verification assessment | |
| Karakatsanis et al. | Data mining approach to monitoring the requirements of the job market: A case study | |
| CN107851097B (en) | Data analysis system, data analysis method, data analysis program, and storage medium | |
| Maleki et al. | A comprehensive literature review of the rank reversal phenomenon in the analytic hierarchy process | |
| Md Saad et al. | Hamming distance method with subjective and objective weights for personnel selection | |
| Boräng et al. | Identifying frames: A comparison of research methods | |
| Ensslin et al. | Seaport-performance tools: an analysis of the international literature | |
| Kostka et al. | Large infrastructure projects in Germany: A cross-sectoral analysis | |
| CN104574110A (en) | Digital credit authentication method | |
| Duarte et al. | An optimization‐based approach for designing attribute acceptance sampling plans | |
| CN107122438A (en) | A kind of judicial case search method and system | |
| Lee et al. | Incremental analysis for generalized TODIM | |
| Gottschalk et al. | Stages of knowledge management technology in the value shop: the case of police investigation performance | |
| CN106228453A (en) | A kind of method and apparatus obtaining user's occupational information | |
| Norlander | New Evidence on Employee Noncompete, No Poach, and No Hire Agreements in the Franchise Sector | |
| CN107220778A (en) | A kind of method, device and the electronic equipment of employee's credit appraisal and application | |
| Toloo et al. | A simplification generalized returns to scale approach for selecting performance measures in data envelopment analysis | |
| Tret'yakova | Russian economic journals indexed in Web of Science: current state and the ways of increasing international visibility | |
| Sari et al. | Application of the Promethee II method for determining road improvement priorities | |
| CN110134866A (en) | Information recommendation method and device | |
| Webster | Principles to guide reliable and ethical research evaluation using metric-based indicators of impact | |
| WO2015118619A1 (en) | Document analysis system, document analysis method, and document analysis program | |
| García et al. | Monitoring credit risk in the social economy sector by means of a binary goal programming model | |
| Abdullah et al. | Mapping crowdfunding research on the web of science database: A bibliometric analysis approach | |
| CN107885706A (en) | A kind of system of data similarity detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: 510033 No. 171, Xin Lian Road, Guangdong, Guangzhou Applicant after: GUANGDONG SCIENCE & TECHNOLOGY INFRASTRUCTURE CENTER Address before: 510033 No. 171, Xin Lian Road, Guangzhou, Guangdong, Yuexiu District Applicant before: Guangdong Computing Center |
|
| CB03 | Change of inventor or designer information |
Inventor after: Luo Liang Inventor after: Lu Zhixing Inventor after: Fang Shaoliang Inventor after: Xu Diwei Inventor after: Cai Jianxin Inventor after: Lin Zhu Inventor before: Luo Liang Inventor before: Lu Zhixing Inventor before: Fang Shaoliang |
|
| COR | Change of bibliographic data | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |