CN104679942A - Construction land bearing efficiency measuring method based on data mining - Google Patents

Abstract

The invention discloses a construction land bearing efficiency measuring method based on data mining. The construction land bearing efficiency measuring method includes steps of S1, mining data to use network open data as main data source and conventional data like demographic census data as complementary, and searching spatial characteristics of construction land bearing efficiency; S2, analyzing data, to be specifically, analyzing addresses in online shopping data and enterprise data; S3, matching data space, subjecting the address analysis data to coordinate conversion after data is collected into position data available for a GIS (geographic information system) platform and correlating the position data with the construction land; S4, realizing regional difference measurement of the construction land bearing efficiency by comprehensive measurement according to multi-element characteristics of the data. By the data mining, the relevant network open data is acquired; by space address matching, social activities of enterprises or microcosmic individual residents are reflected and the problem about measurement of the construction land bearing efficiency is solved.

Description

A kind of construction land load-carrying efficiency Measurement Method based on data mining

Technical field

The present invention relates to the research field of construction land load-carrying efficiency, particularly a kind of construction land load-carrying efficiency Measurement Method based on data mining.

Background technology

In the correlative study of construction land load-carrying efficiency, scholar is had to remove to study relation (the Nazneen Ferdous of land development intensity and land owner with orderly response model in the recent period abroad, 2013), for the diversified development (Yin of dwelling activity, el al, 2011) involved by also having.At home, have scholar to propose the necessity (Gu Xiang etc., 2006) of land use intensively very early, Zeng Yong etc. (2004) also possess some special knowledge with mating of population size to Land_use change.Many scholars economically set out research land utilization efficiency, comprise and set up total factor land utilization efficiency theoretical model (Zhao Xiaobo, 2013), DEA process is used to carry out model investigation (Yuan Lei etc., 2009), stone is recalled Shao (2013) and is also utilized the Land_use change of Information Entropy to Guangzhou College City area to make comprehensive evaluation, also has and utilizes analytical hierarchy process to make the exploration (Shang Tiancheng etc., 2009) of evaluation on Information Entropy basis.But comprehensive bearing capacity of the ground is estimated until in recent years just there is scholar to relate to (Zhu Xiuzhi, 2013).Meanwhile, domestic construction land evolution Feature and impact are also in the middle of research (Li Chang, 2013).But existing research is many based on regional economy aspect, be bad at the discussion in space aspects, the drawback of this kind of research mode based on economy is to overemphasize abstract regional extent aspect, and due to Data Update sluggishness, the assessment of efficiency cannot be carried out, to precision of analysis with ageingly have certain influence to concrete plot.

The progress of infotech accelerates the space-time exchange of knowledge, technology, the talent, fund etc., fluid space becomes the main carriers (Castells of region, city and resident's activity, 1996), have scholar to propose in early days high aggregation body that city is crowd activity, the behavior of people in city occupies more leading factor (I.Gordon to urban study, 2008), and the behavioural characteristic of Social Individual is exactly have ignored with the regional construction land output efficiency analysis that statistical yearbook data are taken as the leading factor, this is also one of defect of existing research.

In recent years, the appearance of large data solves the data shortage problem existed over a long period of time to a certain extent, and the quantitative test that part cannot be carried out becomes feasible (Delyser, Sui, 2013).Also comparatively quantum jump has been had estimating in urban activity, all combined closely with informationalized digital space (Wang Jingyuan etc. 2014) in the aspect such as economy, culture, traffic, amusement in city, the data research of crowd activity and space distribution becomes possibility by social tool such as microbloggings.The research of current application large Data Mining resident trip and traffic is relatively many, and how to carry out studying (Becker, 2011 from number of ways acquisition data such as telecom operators, social network sites, taxi and public smart cards; Kang, 2012; Mark, 2011; Naaman, 2012; Long Ying, Zhang Yu, Cui Chengyin, 2012; Batty, 2013).Emerging in large numbers of correlative study achievement, for the urban and rural planning discipline development of the new period provides fruitful technology guide.Due to the real-time that internet data upgrades, and accurately can locate the geographic position of user, concrete block can be contracted to by the scope of crowd activity's data research, thus, the large data analysis of soil load-carrying efficiency is more effective, avoids the problem that general statistics cannot differentiate concrete ground utilization.

In reality, domestic planning field is bad to estimate the load-carrying efficiency of construction land always, depend on the modes such as economic statistics yearbook data more, although the research mode of this thick bore can differentiate the quality of city built environment under macro-scale, but be limited to the finiteness of space scale, the space characteristics of differentiation soil load-carrying efficiency that cannot be more careful.On the other hand, construction land load-carrying efficiency can not the activity of desocialization's individuality, although the load-carrying efficiency by means of statistical yearbook data is estimated can calculate overall city development degree, but individual space operation is easily ignored in this structural estimating, even and formed between the daily behavior of civic and significantly rupture.

Summary of the invention

Fundamental purpose of the present invention is that the shortcoming overcoming prior art is with not enough, provides the construction land load-carrying efficiency Measurement Method based on data mining.

In order to achieve the above object, the present invention is by the following technical solutions:

Based on a construction land load-carrying efficiency Measurement Method for data mining, comprise the steps:

S1, data mining, originate using network data as general data, and with traditional data such as censuses as a supplement, by the excavation of Activities data and space orientation, explores the space characteristics of construction land load-carrying efficiency; Described network data comprises dwelling activity data, employment activity data, Recreational activities data, social network sites and shopping at network consumption data;

S2, Data Analysis, carry out address resolution for the address in shopping at network data and business data, by the service function of Geocoding API under the Web API page in Baidu LBS open platform, the text message of address is converted to latitude coordinates;

S3, spatial match, upon completing data collection, the address resolution data obtained by preceding method carry out coordinate conversion, are converted into the operable position data of GIS platform, and carry out associating with construction land and mate;

S4, on the basis of data mining, parsing and space dimensionality reduction, based on the diverse characteristics of data, the area position difference being realized construction land load-carrying efficiency by Synthetic Measurement is estimated, and the space specifically comprising single factor data is estimated with the space of integrated data superimposed, is specially:

The space of S4.1, single factor data is estimated, inhabitation, enterprise, the distribution density of Activities of strolling about or have a rest are estimated, utilize open platform, the multidimensional data of enterprises and institutions' register carries out address resolution, obtain the Location distribution of Activities, and in conjunction with the Density Distribution situation of construction land and population, single key element loading strength of reflection urban whole aspect;

The space of S4.2, integrated data is superimposed, in every index factor system, to all kinds of desired value carry out space grating format process after, comprehensive Information Entropy and analytical hierarchy process, agriculture products weight, subjective and objective aggregative weighted calculating is carried out to single key element loading strength, forms the Comprehensive Assessment of construction land load-carrying efficiency thus, and space differentiation is carried out to high level, low value region.

Preferably, in step S1, described social network data comprises the data of Sina's microblogging, for the data acquisition of Sina's microblogging, mainly obtained text message and the geographic coordinate information of its issuing microblog by the microblogging interface in microblogging API service and geography information interface, be specially:

S1.1, fill in web application information according to related request, obtain App Key and App Secret;

S1.2, log in the open platform of Sina's microblogging, enter data grabber interface, parameters in optimum configurations region, click calling interface, the result returned is asked in display;

S1.3, enter " 2/place/nearby_timeline " sub-interface in geography information interface, request URL is write according to required parameter, the central point determining to capture, the beginning and ending time of search radius and search, Python files in batch obtains search content, described search content comprises: the temporal information of issuing microblog, geographic coordinate information, issue content of text, user id, distance center point distance, sex and user's location message, and the concrete steps being obtained search content by Python files in batch are as follows:

S1.3.1, instantiation APIClient class object, send Auth authentication request to user, and the client user account password of having preserved logs in, and simulation is agreed to authorize;

S1.3.2, obtain user agree to authorize after URL:

YOUR_REGISTERED_REDIRECT_URI/? code=CODE, monitors HTTP process, intercepts the value of code in readjustment webpage;

S1.3.3, submission code, to API service device, use the OAuth2.0Access Token obtained to call API;

S1.3.4, obtain comprising the number of registering of POI point, the JSON formatted data with the information such as microblogging, user profile in geographic position by calling the interfaces such as place/pois/users, place/pois/tips, place/poi_timeline;

The data of S1.3.5, parsing Json form, are kept at this locality.

Preferably, described network data also comprises image data, for the crawl of image data, undertaken by panoramio website, by the Web API service function under the Geocoding API page in panoramio open platform, according to its parameter request, write the URL sending http request, and by LocoySpider software, the data that Batch sending and reception http request return.

Preferably, in step S2, the concrete grammar of Data Analysis is:

S2.1, acquisition api interface key;

The key that S2.2, use obtain, according to its parameter request, writes the URL sending http request;

S2.3, by LocoySpider software, Batch sending and receive the data that return of http request, completes the address resolution of text message thus.

Preferably, in step S4.1, in single key element of construction land load-carrying efficiency is estimated, start with from the intensity distributions of three major types basic activity, first the quantity of the atomization data of the dwelling activity in grid, Recreational activities is gathered, business activity is then gathered by the sales volume, and calculates the activity intensity r of kth item key element in space cell i _ik, and the absolute intensity data of Activities is normalized, that is:

$r_{\mathrm{ik}}=\frac{\frac{\mathrm{\Σ}{N}_{\mathrm{ik}}}{S_i}}{\max \left(\frac{\mathrm{\Σ}{N}_{\mathrm{ik}}}{S_i}\right)}---\left(1\right)$

Wherein S _ibe the construction land area of the i-th grid space unit, N _ibe the i-th grid k item activity quantity and; Arcscene software is used to be shown by its space distribution three dimensional stress again.

Preferably, in step S4.2, in Synthetic Measurement, by four classes subitem key element activity datas and demographic data space superimposed, Comprehensive analytic hierarchy process (AHP) and entropy assessment, be weighted five class indexs, form the Comprehensive Assessment of construction land load-carrying efficiency, the comprehensive evaluation model of construction land efficiency is:

$y_i=\underset{k=1}{\overset{5}{\mathrm{\Σ}}}{\mathrm{\α}}_k{y}_{\mathrm{ik}}---\left(2\right)$

For the comprehensive weight α of all kinds of key element activity _k, adopt combination weights method, note p _k, q _krepresent k respectively

Item key element composes power by analytic hierarchy process AHP, entropy assessment composes the weight weighed and obtain,

α _k＝c ₁ ^*p _k+c ₂ ^*q _k(3)。

Preferably, in S4.2, determine the weight p of each key element according to analytic hierarchy process AHP _k, be specially:

S4.2.1, set up hierarchy Model, Judgement Matricies A;

$A=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ ...& ...& ...& ...\\ a_{k1}& a_{k2}& ...& a_{\mathrm{kn}}\end{array}\right]---\left(4\right)$

S4.2.2, determine the importance ranking of every key element activity by sequence between two; To judgment matrix A, calculate and meet AW=λ _maxw, wherein λ _maxfor the Maximum characteristic root of A, W is for corresponding to λ _maxregular proper vector, the component W of W _kbe exactly the weighted value of the single sequence of corresponding element, application root method solves normalization characteristic vector sum eigenwert; Be specially:

First the n th Root of the product of each rower degree of judgment matrix is calculated

$\stackrel{\‾}{W_k}=\sqrt[n]{M_k}---\left(5\right)$

$M_k=\underset{j=1}{\overset{n}{\mathrm{\Π}}}{a}_{\mathrm{kj}}---\left(6\right)$

M _kfor the value of each row element in judgment matrix;

Root vector is normalized, obtains a kth component of characteristic vector W

$W_k=\frac{\stackrel{\‾}{W_k}}{{\mathrm{\Σ}}_{k=1}^5\stackrel{\‾}{W_k}}---\left(7\right)$

Finally calculate the Maximum characteristic root of judgment matrix

${\mathrm{\λ}}_{\max }=\underset{k=1}{\overset{5}{\mathrm{\Σ}}}\frac{{\left(\mathrm{AW}\right)}_k}{{\mathrm{nW}}_k}---\left(8\right)$

(AW) _kfor a kth component of vectorial AW;

S4.2.3, again through consistency check, finally determine the sequence of every key element activity;

The consistance of test and judge matrix, calculates its coincident indicator:

CI＝(λ _max-k)/(k-1) (9)

And CI and Aver-age Random Consistency Index RI is compared, be designated as CR, as CR=CI/RI<0.10, think that judgment matrix has gratifying consistance;

S4.2.4, determine the weight p of the every key element activity of analytical hierarchy process _k;

$p_k=\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}{a}_j{b}_k^j---\left(10\right)$

P in formula _kfor indicator layer weight, a _jfor each key element of rule layer is relative to the weight of destination layer, for index is the weight of each factor relative to rule layer;

Total sequence consistance formula:

$\mathrm{CR}=\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}\left(a_j{\mathrm{CI}}_j\right)/\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}\left(a_j{\mathrm{RI}}_j\right)---\left(11\right)$

As CR<0.10, think that total sequence has satisfied consistance, namely obtain the weight of each key element.

Preferably, q is calculated according to entropy assessment _kdetermine the objective weight of all kinds of activity, concrete grammar is:

S4.2.5, in units of the street of town, calculate each unit permanent resident population density;

S4.2.6, according to Information Entropy determination weight coefficient, then to have

${\mathrm{\&beta;}}_{\mathrm{ik}}=\frac{r_{\mathrm{ik}}-r_{\min }}{r_{\max }-r_{\min }}---\left(12\right)$

Wherein r _max, r _minrepresent the maximal value of k item key element in all grid cells, minimum value respectively;

S4.2.7, calculate the entropy of every factors evaluation index, first define

$f_{\mathrm{ik}}={\mathrm{\&beta;}}_{\mathrm{ik}}/\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&beta;}}_{\mathrm{ik}}---\left(13\right)$

In formula: i=1,2 ..., m; K=1,2,3,4,5

Entropy according to the every factors evaluation index of the concept definition of entropy is:

$H_k=-\frac{1}{\ln \left(m\right)}\&CenterDot;\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{f}_{\mathrm{ik}}\&CenterDot;\ln \left(f_{\mathrm{ik}}\right)---\left(14\right)$

Due to f _ikwhen=0, lnf _ikmeaningless, therefore the method for reference Meng Xianmeng (2009), by f _ikrevise, then:

$f_{\mathrm{ik}}=(1+{\mathrm{\&beta;}}_{\mathrm{ik}})/\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}(1+{\mathrm{\&beta;}}_{\mathrm{ik}})---\left(15\right)$

S4.2.8, the entropy utilizing revised measurement index to calculate each key element are weighed:

$q_k=H_k/\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{H}_k---\left(16\right)$

Above-mentioned formula meets

Preferably, in step S4, the Synthetic Measurement of load-carrying efficiency is specially:

S5.1, calculate the comprehensive load-carrying efficiency of each grid cell;

$y_i=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{y}_{\mathrm{ik}}=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}({c_1}^{*}{p}_k+{c_2}^{*}{q}_k)y_{\mathrm{ik}}---\left(17\right)$

S5.2, by Arcscene software by its space distribution three dimensional stress show, obtain the comprehensive evaluation of construction land load-carrying efficiency thus;

Data, according to the composite score finally determined, by the space cell that differential technique search load-carrying efficiency is lower, and are normalized, calculate the low value score of each space cell by S5.3, calculation low value score

${\mathrm{\&mu;}}_i=\frac{1-y_i}{\max (1-y_i)}---\left(18\right)$

S5.4, calculating mean value μ, calculate the arithmetic mean of construction land load-carrying efficiency in all grid cells, and effective grid cell number is m,

$\mathrm{\&mu;}=\frac{{\mathrm{\&mu;}}_i}{m}---\left(19\right)$

S5.5, calculating standard deviation δ, the low value calculated according to upper two steps and mean value, calculate the standard deviation δ of all grid cell construction land load-carrying efficiencies;

$\mathrm{\&delta;}=\sqrt{\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{({\mathrm{\&mu;}}_i-\mathrm{\&mu;})}^2}---\left(20\right)$

S5.6, low value area judging, the multiple departing from standard deviation by each cell-average difference finds out load-carrying efficiency depression, then have

$Q_i=\frac{{\mathrm{\&mu;}}_i-\mathrm{\&mu;}}{\mathrm{\&delta;}}---\left(21\right)$

S5.7, calculate each town street grid number N _t, with town street for unit, select large principle by GIS software and calculate to belong to area the grid cell number N that each town street comprises _t;

S5.8, in units of the street of town identification effect depression, calculate the mean value that each town street construction land carrying depression unit number departs from whole city's standard deviation multiple;

$\stackrel{\&OverBar;}{Q_t}=\frac{\mathrm{\&Sigma;}{Q}_{\mathrm{it}}}{N_t}---\left(22\right)$

Judge whether the construction land in each town street belongs to the region, depression of load-carrying efficiency thus.

Compared with prior art, tool has the following advantages and beneficial effect in the present invention:

1, the larger space cell such as existing construction land assessment technique Yi Shi, county, district is measurement object, is difficult to differentiate the ground utilization in concrete plot.The present invention attempts the mode adopting data mining, obtains relevant network data, based on the social activities that enterprise or resident etc. are individual, in the mode of spatial match association, solves the evaluation problem of microcosmic point construction land load-carrying efficiency.

2, prior art depends on the statistical data of national economy, and correlation discriminating result depends on existing statistics.The present invention utilizes network opening data, and business directory data also can be obtained by industrial and commercial bureau, and the process of address resolution is also utilize existing network opening platform, and this mode effectively overcomes in existing analytical technology estimates Data Collection disadvantage not easily.

Accompanying drawing explanation

Fig. 1 is process flow diagram of the present invention;

Fig. 2 (a)-Fig. 2 (d) is respectively the movable load-carrying efficiency schematic diagram that net purchase data, enterprise's sales volume, microblog data and image data table characterize;

Fig. 3 is the comprehensive evaluation schematic diagram of construction land load-carrying efficiency in the present embodiment;

Fig. 4 is Dongguan construction land load-carrying efficiency low value region signal in the present embodiment.

Embodiment

Below in conjunction with embodiment and accompanying drawing, the present invention is described in further detail, but embodiments of the present invention are not limited thereto.

Embodiment

As shown in Figure 1, the present embodiment, based on the construction land load-carrying efficiency Measurement Method of data mining, comprises the steps:

(1), data mining, this method using network data as main Data Source, and with traditional data such as censuses as a supplement.Dwelling activity, data from shopping at network consumption data, carries out address resolution by the information of receiving, and obtains position, living space; Employment activity data derives from enterprises and institutions' register in 2014; Recreational activities is data from social networks Sina microblogging open platform (http://open.weibo.com), and with panoramio website under Google house flag (http://www.panoramio.com) image data as a supplement, obtain the geospatial information of its user's issuing microblog and uploading pictures.By to the excavation of Activities data and space orientation, explore the space characteristics of construction land load-carrying efficiency.The time that wherein social network sites and shopping at network consumption data obtain is the 6-7 month in 2014, image data is all information that 2006-2014 comprises, employment data source is Yi Mei database, relevant geographical spatial coordinated information adopts Baidu's open platform to carry out space orientation, and imports generalized information system after coordinates correction.

(1-1), for the data acquisition of Sina's microblogging, text message and the geographic coordinate information of its issuing microblog is mainly obtained by the microblogging interface in microblogging API service and geography information interface, concrete steps are:

(1-1-1) fill in web application information according to related request, obtain App Key and App Secret; It should be noted that, although be not directly involved in above-mentioned two data in subsequent step, it is the necessary step of data grabber.

(1-1-2) log in the open platform of Sina's microblogging, enter data grabber interface, input parameter in the frame in optimum configurations region, click calling interface, can see in the green frame in the right side and returning results.Required parameter illustrates as shown in table 1.

Table 1API test interface parameter declaration

(1-1-3) " 2/place/nearby_timeline " sub-interface in geography information interface is entered, request URL is write according to association requests parameter, the central point determining to capture, the beginning and ending time of search radius and search, the Python files in batch using author to write obtains related content.Obtainable content comprises: the temporal information, geographic coordinate information, issue content of text, user id, distance center point distance, sex, user location etc. of issuing microblog; The concrete steps being obtained search content by Python files in batch are as follows:

(1-1-3-1) instantiation APIClient class object, sends Auth authentication request to user, and the client user account password of having preserved logs in, and simulation is agreed to authorize;

(1-1-3-2) obtain user agree to authorize after URL:YOUR_REGISTERED_REDIRECT_URI/? code=CODE, monitors HTTP process, intercepts the value of code in readjustment webpage;

(1-1-3-3) submit to code to API service device, use the OAuth2.0Access Token obtained to call API;

(1-1-3-4) obtain comprising the number of registering of POI point, the JSON formatted data with the information such as microblogging, user profile in geographic position by calling the interfaces such as place/pois/users, place/pois/tips, place/poi_timeline;

(1-1-3-5) resolve the data of Json form, be kept at this locality.

(1-2) for the crawl of image data, carry out mainly through panoramio website (http://www.panoramio.com), by the Web API service function under the Geocoding API page in panoramio open platform, according to its parameter request, write the URL sending http request, and by LocoySpider software, the data that Batch sending and reception http request return.

(2) Data Analysis, carries out address resolution for the address in shopping at network data and business data.By the service function of Geocoding API under the Web API page in Baidu LBS open platform, the text message of address is converted to latitude coordinates.Concrete steps are:

(2.1) api interface key is obtained;

(2.2) use the key obtained, according to its parameter request, write the URL sending http request; As shown in table 2;

Table 2

The exclusive required parameter of geocoding:

(2.3) by LocoySpider software, the data that Batch sending and reception http request return, complete the address resolution of text message thus.

(3) space dimensionality reduction, upon completing data collection, the address resolution data obtained by preceding method carry out coordinate conversion, are converted into the operable position data of GIS platform; Divide the space cell of construction land efficiency analysis.Control according to the population size about urban community 3-5 ten thousand in " Code of urban Residential areas planning m Design (2002) ", rasterizing data being carried out to 2km × 2km merges, and is converted into the spatial data that GIS software can be analyzed.

In case subjects, choose Dongguan demonstrate as an example, development scale and the economic level of Dongguan accumulation over more than 30 years of reform and opening-up also ensure that space scale needed for data mining and data scale.

(4) on the basis of data mining, parsing and space dimensionality reduction, based on the diverse characteristics of data, the area position difference being realized construction land load-carrying efficiency by Synthetic Measurement is estimated, and the space specifically comprising single factor data is estimated with the space of integrated data superimposed, is specially:

(4.1) space of single factor data is estimated, inhabitation, enterprise, the distribution density of Activities of strolling about or have a rest are estimated, utilize open platform, the multidimensional data of enterprises and institutions' register carries out address resolution, obtain the Location distribution of Activities, and in conjunction with the Density Distribution situation of construction land and population, single key element loading strength of reflection urban whole aspect;

(4.2) space of integrated data is superimposed, in every index factor system, to all kinds of desired value carry out space grating format process after, comprehensive Information Entropy and analytical hierarchy process, agriculture products weight, subjective and objective aggregative weighted calculating is carried out to single key element loading strength, forms the Comprehensive Assessment of construction land load-carrying efficiency thus, and space differentiation is carried out to high level, low value region.

In above-mentioned steps (4.1), in single key element of construction land load-carrying efficiency is estimated, start with from the intensity distributions of three major types basic activity, first the quantity of the atomization data of the dwelling activity in grid, Recreational activities is gathered, business activity is then gathered by the sales volume, and calculates the activity intensity r of kth item key element in space cell i _ik, and the absolute intensity data of Activities is normalized, that is:

$r_{\mathrm{ik}}=\frac{\frac{\mathrm{\&Sigma;}{N}_{\mathrm{ik}}}{S_i}}{\max \left(\frac{\mathrm{\&Sigma;}{N}_{\mathrm{ik}}}{S_i}\right)}---\left(1\right)$

Wherein S _ibe the construction land area of the i-th grid space unit, N _ibe the i-th grid k item activity quantity and; Arcscene software is used to be shown, as shown in Fig. 2 (a)-Fig. 2 (d) by its space distribution three dimensional stress again.

Calculate the dwelling activity load-carrying efficiency distribution plan of shopping at network data.Wherein the dwelling activity intensity in Guan Cheng street is the highest.In administrative region of a city aspect, the dwelling activity of Dongguan also presents multicenter Distribution Pattern, but other town street dwelling activity intensity except Guan Cheng street are more close

Calculate the business activity load-carrying efficiency distribution plan of enterprise's sales volume.The polarization characteristic of Dongguan is comparatively obvious, and the business activity of Guan Cheng-Dongcheng, two places, Chang'an is with the obvious advantage compared with other town streets, and in addition, near the Ma Yong town in Guangzhou, its business activity intensity is also relatively high, and Chang Ping, the business activity of camphorwood first-class ground in east side are more weak.

Calculate the Recreational activities intensity distribution of microblog data.As shown in the figure, relative to inhabitation and business activity, Recreational activities distribution is more extensive, and the Recreational activities in the street such as Wan Jiang, southern city also embodies to some extent, but mostly all concentrates on down town and neighboring area, and the trend of outwards being successively decreased by center is obvious.

Comparatively speaking, the Recreational activities that image data characterizes is continuous trend then more obviously, and show that the activities such as resident's Recreational activities scope is comparatively lived, employment are larger, polarization characteristic is not obvious.The Recreational activities on Dongcheng, tea hill, the first-class ground of camphorwood all has upper zone, and mass activity intensity is higher.Wide dark riverine expressway does not form obvious unbroken region along the line.

In above-mentioned steps (4.2), as shown in Figure 3, in Synthetic Measurement, by four classes subitem key element activity datas and demographic data space superimposed, Comprehensive analytic hierarchy process (AHP) and entropy assessment, be weighted five class indexs, form the Comprehensive Assessment of construction land load-carrying efficiency, the comprehensive evaluation model of construction land efficiency is:

$y_i=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{y}_{\mathrm{ik}}---\left(2\right)$

For the comprehensive weight α of all kinds of key element activity _k, adopt combination weights method, note p _k, q _krepresent k respectively

Item key element composes power by analytic hierarchy process AHP, entropy assessment composes the weight weighed and obtain,

α _k＝c ₁ ^*p _k+c ₂ ^*q _k(3)。

Then, the weight p of each key element is determined according to analytic hierarchy process AHP _k, be specially:

(4.2.1) hierarchy Model is set up, Judgement Matricies A;

$A=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ ...& ...& ...& ...\\ a_{k1}& a_{k2}& ...& a_{\mathrm{kn}}\end{array}\right]---\left(4\right)$

(4.2.2) importance ranking of every key element activity is determined by sequence between two; To judgment matrix A, calculate and meet AW=λ _maxw, wherein λ _maxfor the Maximum characteristic root of A, W is for corresponding to λ _maxregular proper vector, the component W of W _kbe exactly the weighted value of the single sequence of corresponding element, application root method solves normalization characteristic vector sum eigenwert; Be specially:

First the n th Root of the product of each rower degree of judgment matrix is calculated

$\stackrel{\&OverBar;}{W_k}=\sqrt[n]{M_k}---\left(5\right)$

$M_k=\underset{j=1}{\overset{n}{\mathrm{\&Pi;}}}{a}_{\mathrm{kj}}---\left(6\right)$

M _kfor the value of each row element in judgment matrix;

Root vector is normalized, obtains a kth component of characteristic vector W

$W_k=\frac{\stackrel{\&OverBar;}{W_k}}{{\mathrm{\&Sigma;}}_{k=1}^5\stackrel{\&OverBar;}{W_k}}---\left(7\right)$

Finally calculate the Maximum characteristic root of judgment matrix

${\mathrm{\&lambda;}}_{\max }=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}\frac{{\left(\mathrm{AW}\right)}_k}{{\mathrm{nW}}_k}---\left(8\right)$

(AW) _kfor a kth component of vectorial AW;

(4.2.3) again through consistency check, the sequence of every key element activity is finally determined;

The consistance of test and judge matrix, calculates its coincident indicator:

CI＝(λ _max-k)/(k-1)(9)

And CI and Aver-age Random Consistency Index RI is compared, be designated as CR, as CR=CI/RI<0.10, think that judgment matrix has gratifying consistance;

(4.2.4) the weight p of the every key element activity of analytical hierarchy process is determined _k;

$p_k=\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}{a}_j{b}_k^j---\left(10\right)$

P in formula _kfor indicator layer weight, a _jfor each key element of rule layer is relative to the weight of destination layer, for index is the weight of each factor relative to rule layer;

Total sequence consistance formula:

$\mathrm{CR}=\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}\left(a_j{\mathrm{CI}}_j\right)/\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}\left(a_j{\mathrm{RI}}_j\right)---\left(11\right)$

As CR<0.10, think that total sequence has satisfied consistance, namely obtain the weight of each key element.

Q is calculated again according to entropy assessment _kdetermine the objective weight of all kinds of activity, concrete grammar is:

(4.2.5) in units of the street of town, each unit permanent resident population density is calculated;

(4.2.6) according to Information Entropy determination weight coefficient, then have

${\mathrm{\&beta;}}_{\mathrm{ik}}=\frac{r_{\mathrm{ik}}-r_{\min }}{r_{\max }-r_{\min }}---\left(12\right)$

Wherein r _max, r _minrepresent the maximal value of k item key element in all grid cells, minimum value respectively;

(4.2.7) entropy of every factors evaluation index is calculated.First define

$f_{\mathrm{ik}}={\mathrm{\&beta;}}_{\mathrm{ik}}/\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&beta;}}_{\mathrm{ik}}---\left(13\right)$

In formula: i=1,2 ..., m; K=1,2,3,4,5,

Entropy according to the every factors evaluation index of the concept definition of entropy is:

$H_k=-\frac{1}{\ln \left(m\right)}\&CenterDot;\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{f}_{\mathrm{ik}}\&CenterDot;\ln \left(f_{\mathrm{ik}}\right)---\left(14\right)$

Due to f _ikwhen=0, lnf _ikmeaningless, therefore the method for reference Meng Xianmeng (2009), by f _ikrevise, then:

$f_{\mathrm{ik}}=(1+{\mathrm{\&beta;}}_{\mathrm{ik}})/\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}(1+{\mathrm{\&beta;}}_{\mathrm{ik}})---\left(15\right)$

(4.2.8) revised measurement index is utilized to calculate the entropy power of each key element:

$q_k=H_k/\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{H}_k---\left(16\right)$

Above-mentioned formula meets

In step S4, the Synthetic Measurement of load-carrying efficiency is specially:

(4.3) by main, that the objective mode combined calculates each construction land grid cell comprehensive load-carrying efficiency;

$y_i=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{y}_{\mathrm{ik}}=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}({c_1}^{*}{p}_k+{c_2}^{*}{q}_k)y_{\mathrm{ik}}---\left(17\right)$

(4.4) by Arcscene software, its space distribution three dimensional stress is shown, obtain the comprehensive evaluation of construction land load-carrying efficiency thus;

(4.5) calculate low value score, according to the composite score finally determined, by the space cell that differential technique search load-carrying efficiency is lower, and data are normalized, calculate the low value score of each space cell

${\mathrm{\&mu;}}_i=\frac{1-y_i}{\max (1-y_i)}---\left(18\right)$

(4.6) calculating mean value μ, calculates the arithmetic mean of construction land load-carrying efficiency in all grid cells, and effective grid cell number is m,

$\mathrm{\&mu;}=\frac{{\mathrm{\&mu;}}_i}{m}---\left(19\right)$

(4.7) calculate standard deviation δ, the low value calculated according to upper two steps and mean value, calculate the standard deviation δ of all grid cell construction land load-carrying efficiencies;

$\mathrm{\&delta;}=\sqrt{\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{({\mathrm{\&mu;}}_i-\mathrm{\&mu;})}^2}---\left(20\right)$

(4.8) low value area judging, the multiple departing from standard deviation by each space cell mean difference finds out load-carrying efficiency depression, then have

$Q_i=\frac{{\mathrm{\&mu;}}_i-\mathrm{\&mu;}}{\mathrm{\&delta;}}---\left(21\right)$

(4.9) each town street grid number N is calculated _t, with town street for unit, select large principle by GIS software and calculate to belong to area the grid cell number N that each town street comprises _t;

(4.10) identification effect depression in units of the street of town, calculates the mean value that construction land carrying depression, each town street unit number departs from whole city's standard deviation multiple;

$\stackrel{\&OverBar;}{Q_t}=\frac{\mathrm{\&Sigma;}{Q}_{\mathrm{it}}}{N_t}---\left(22\right)$

Judge whether the construction land in each town street belongs to the region, depression of load-carrying efficiency thus.

Synthesizing map 4 content can be found out, the region that load-carrying efficiency is higher still concentrates on downtown area, and east side, Dongguan is near the town street in Huizhou, and the load-carrying efficiency of its entirety is all lower.The construction land load-carrying efficiency in central scroll town is minimum, the Jie Weiqishi town, town of next, Xie Gang town, camphorwood head town.And the town street of load-carrying efficiency on average level is also more, the load-carrying efficiency in the Humen, Dongcheng, Guan Cheng and Chang'an etc. is all higher.The highest region of load-carrying efficiency is in Guan Cheng-region, Dongcheng, and very important, and the town streets such as the fiber crops of adjoining Guangzhou are gushed, central scroll, while its construction land is expanded on a large scale, load-carrying efficiency has much room for improvement.

Above-described embodiment is the present invention's preferably embodiment; but embodiments of the present invention are not restricted to the described embodiments; change, the modification done under other any does not deviate from Spirit Essence of the present invention and principle, substitute, combine, simplify; all should be the substitute mode of equivalence, be included within protection scope of the present invention.

Claims (9)

1., based on a construction land load-carrying efficiency Measurement Method for data mining, it is characterized in that, comprise the steps:

S1, data mining, originate using network data as general data, and with traditional data such as censuses as a supplement, by the excavation of Activities data and space orientation, explores the space characteristics of construction land load-carrying efficiency; Described network data comprises dwelling activity data, employment activity data, Recreational activities data, social network sites and shopping at network consumption data;

S2, Data Analysis, carry out address resolution for the address in shopping at network data and business data, by the service function of Geocoding API under the Web API page in Baidu LBS open platform, the text message of address is converted to latitude coordinates;

S3, spatial match, upon completing data collection, the address resolution data obtained by preceding method carry out coordinate conversion, are converted into the operable position data of GIS platform, and carry out associating with construction land and mate;

S4, on the basis of data mining, parsing and space dimensionality reduction, based on the diverse characteristics of data, the area position difference being realized construction land load-carrying efficiency by Synthetic Measurement is estimated, and the space specifically comprising single factor data is estimated with the space of integrated data superimposed, is specially:

The space of S4.1, single factor data is estimated, inhabitation, enterprise, the distribution density of Activities of strolling about or have a rest are estimated, utilize open platform, the multidimensional data of enterprises and institutions' register carries out address resolution, obtain the Location distribution of Activities, and in conjunction with the Density Distribution situation of construction land and population, single key element loading strength of reflection urban whole aspect;

The space of S4.2, integrated data is superimposed, in every index factor system, to all kinds of desired value carry out space grating format process after, Comprehensive analytic hierarchy process AHP and Information Entropy, agriculture products weight, subjective and objective aggregative weighted calculating is carried out to single key element loading strength, forms the Comprehensive Assessment of construction land load-carrying efficiency thus, and space differentiation is carried out to high level, low value region.

2. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 1, it is characterized in that, in step S1, described social network data comprises the data of Sina's microblogging, for the data acquisition of Sina's microblogging, mainly obtained text message and the geographic coordinate information of its issuing microblog by the microblogging interface in microblogging API service and geography information interface, be specially:

S1.1, fill in web application information according to related request, obtain App Key and App Secret;

S1.2, log in the open platform of Sina's microblogging, enter data grabber interface, parameters in optimum configurations region, click calling interface, the result returned is asked in display;

S1.3, enter " 2/place/nearby_timeline " sub-interface in geography information interface, request URL is write according to required parameter, the central point determining to capture, the beginning and ending time of search radius and search, Python files in batch obtains search content, described search content comprises: the temporal information of issuing microblog, geographic coordinate information, issue content of text, user id, distance center point distance, sex and user's location message, and the concrete steps being obtained search content by Python files in batch are as follows:

S1.3.1, instantiation APIClient class object, send Auth authentication request to user, and the client user account password of having preserved logs in, and simulation is agreed to authorize;

S1.3.2, obtain user agree to authorize after URL:

YOUR_REGISTERED_REDIRECT_URI/? code=CODE, monitors HTTP process, intercepts the value of code in readjustment webpage;

S1.3.3, submission code, to API service device, use the OAuth2.0 Access Token obtained to call API;

S1.3.4, obtain comprising the number of registering of POI point, the JSON formatted data with the information such as microblogging, user profile in geographic position by calling the interfaces such as place/pois/users, place/pois/tips, place/poi_timeline;

The data of S1.3.5, parsing Json form, are kept at this locality.

3. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 1, it is characterized in that, described network data also comprises image data, for the crawl of image data, be undertaken by panoramio website, by the Web API service function under the Geocoding API page in panoramio open platform, according to its parameter request, write the URL sending http request, and by LocoySpider software, the data that Batch sending and reception http request return.

4. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 1, is characterized in that, in step S2, the concrete grammar of Data Analysis is:

S2.1, acquisition api interface key;

The key that S2.2, use obtain, according to its parameter request, writes the URL sending http request;

S2.3, by LocoySpider software, Batch sending and receive the data that return of http request, completes the address resolution of text message thus.

5. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 1, it is characterized in that, in step S4.1, in single key element of construction land load-carrying efficiency is estimated, start with from the intensity distributions of three major types basic activity, first gather the quantity of the atomization data of the dwelling activity in grid, Recreational activities, business activity is then gathered by the sales volume, and calculates the activity intensity r of kth item key element in space cell i _ik, and the absolute intensity data of Activities is normalized, that is:

$r_{\mathrm{ik}}=\frac{\frac{\mathrm{\&Sigma;}{N}_{\mathrm{ik}}}{S_i}}{\max \left(\frac{\mathrm{\&Sigma;}{N}_{\mathrm{ik}}}{S_i}\right)}---\left(1\right)$

Wherein S _ibe the construction land area of the i-th grid space unit, N _ibe the i-th grid k item activity quantity and; Arcscene software is used to be shown by its space distribution three dimensional stress again.

6. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 1, it is characterized in that, in step S4.2, in Synthetic Measurement, by four classes subitem key element activity datas and demographic data space superimposed, Comprehensive analytic hierarchy process AHP and entropy assessment, be weighted five class indexs, form the Comprehensive Assessment of construction land load-carrying efficiency, the comprehensive evaluation model of construction land efficiency is:

$y_i=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{y}_{\mathrm{ik}}---\left(2\right)$

For the comprehensive weight α of all kinds of key element activity _k, adopt combination weights method, note p _k, q _krepresent k respectively

Item key element composes power by analytic hierarchy process AHP, entropy assessment composes the weight weighed and obtain,

α _k＝c ₁ ^*p _k+c ₂ ^*q _k(3)。

7. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 6, is characterized in that, in S4.2, determine the weight p of each key element according to analytic hierarchy process AHP _k, be specially:

S4.2.1, set up hierarchy Model, Judgement Matricies A;

$A=\left[\begin{array}{cccc}{a}_{11}& a_{12}& ...& a_{1n}\\ ...& ...& ...& ...\\ a_{k1}& a_{k2}& ...& a_{\mathrm{kn}}\end{array}\right]---\left(4\right)$

S4.2.2, determine the importance ranking of every key element activity by sequence between two; To judgment matrix A, calculate and meet AW=λ _maxw, wherein λ _maxfor the Maximum characteristic root of A, W is for corresponding to λ _maxregular proper vector, the component W of W _kbe exactly the weighted value of the single sequence of corresponding element, application root method solves normalization characteristic vector sum eigenwert; Be specially:

First the n th Root of the product of each rower degree of judgment matrix is calculated

$\stackrel{\&OverBar;}{W_k}=\sqrt[n]{M_k}---\left(5\right)$

$M_k=\underset{j=1}{\overset{n}{\mathrm{\&Pi;}}}{a}_{\mathrm{kj}}---\left(6\right)$

M _kfor the value of each row element in judgment matrix;

Root vector is normalized, obtains a kth component of characteristic vector W

$W_k=\frac{\stackrel{\&OverBar;}{W_k}}{{\mathrm{\&Sigma;}}_{k=1}^5\stackrel{\&OverBar;}{W_k}}---\left(7\right)$

Finally calculate the Maximum characteristic root of judgment matrix

${\mathrm{\&lambda;}}_{\max }=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}\frac{{\left(\mathrm{AW}\right)}_k}{{\mathrm{nW}}_k}---\left(8\right)$

(AW) _kfor a kth component of vectorial AW;

S4.2.3, again through consistency check, finally determine the sequence of every key element activity;

The consistance of test and judge matrix, calculates its coincident indicator:

CI＝(λ _max-k)/(k-1) (9)

And CI and Aver-age Random Consistency Index RI is compared, be designated as CR, as CR=CI/RI<0.10, think that judgment matrix has gratifying consistance;

S4.2.4, determine the weight p of the every key element activity of analytical hierarchy process _k;

$p_k=\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}{a}_j{b}_k^j---\left(10\right)$

P in formula _kfor indicator layer weight, a _jfor each key element of rule layer is relative to the weight of destination layer, for index is the weight of each factor relative to rule layer;

Total sequence consistance formula:

$\mathrm{CR}=\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}(a_j,{\mathrm{CI}}_j)/\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}\left(a_j{\mathrm{RI}}_j\right)---\left(11\right)$

As CR<0.10, think that total sequence has satisfied consistance, namely obtain the weight of each key element.

8. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 6, is characterized in that, calculates q according to entropy assessment _kdetermine the objective weight of all kinds of activity, concrete grammar is:

S4.2.5, in units of the street of town, calculate each unit permanent resident population density;

S4.2.6, according to Information Entropy determination weight coefficient, then to have

${\mathrm{\&beta;}}_{\mathrm{ik}}=\frac{r_{\mathrm{ik}}-r_{\min }}{r_{\max }-r_{\min }}---\left(12\right)$

Wherein r _max, r _minrepresent the maximal value of k item key element in all grid cells, minimum value respectively;

S4.2.7, calculate the entropy of every factors evaluation index, first define

$f_{\mathrm{ik}}={\mathrm{\&beta;}}_{\mathrm{ik}}/\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&beta;}}_{\mathrm{ik}}---\left(13\right)$

In formula: i=1,2 ..., m; K=1,2,3,4,5

Entropy according to the every factors evaluation index of the concept definition of entropy is:

$H_k=-\frac{1}{\ln \left(m\right)}\&CenterDot;\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{f}_{\mathrm{ik}}\&CenterDot;\ln \left(f_{\mathrm{ik}}\right)---\left(14\right)$

Due to f _ikwhen=0, lnf _ikmeaningless, therefore the method for reference Meng Xianmeng (2009), by f _ikrevise, then:

$f_{\mathrm{ik}}=(1+{\mathrm{\&beta;}}_{\mathrm{ik}})/\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}(1+{\mathrm{\&beta;}}_{\mathrm{ik}})---\left(15\right)$

S4.2.8, the entropy utilizing revised measurement index to calculate each key element are weighed:

$q_k=H_k/\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{H}_k---\left(16\right)$

Above-mentioned formula meets

9. the construction land load-carrying efficiency Measurement Method based on data mining according to claim 1, it is characterized in that, in step S4, the Synthetic Measurement of load-carrying efficiency is specially:

S5.1, calculate the comprehensive load-carrying efficiency of each grid cell;

$y_i=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_k{y}_{\mathrm{ik}}=\underset{k=1}{\overset{5}{\mathrm{\&Sigma;}}}({c_1}^{*}{p}_k+{c_2}^{*}{q}_k)y_{\mathrm{ik}}---\left(17\right)$

S5.2, by Arcscene software by its space distribution three dimensional stress show, obtain the comprehensive evaluation of construction land load-carrying efficiency thus;

Data, according to the composite score finally determined, by the space cell that differential technique search load-carrying efficiency is lower, and are normalized, calculate the low value score of each space cell by S5.3, calculation low value score

${\mathrm{\&mu;}}_i=\frac{{1-y}_i}{\max \left({1-y}_i\right)}---\left(18\right)$

S5.4, calculating mean value μ, calculate the arithmetic mean of construction land load-carrying efficiency in all grid cells, and effective grid cell number is m,

$\mathrm{\&mu;}=\frac{{\mathrm{\&mu;}}_i}{m}---\left(19\right)$

S5.5, calculating standard deviation δ, the low value calculated according to upper two steps and mean value, calculate the standard deviation δ of all grid cell construction land load-carrying efficiencies;

$\mathrm{\&delta;}=\sqrt{\frac{1}{m}\underset{i=1}{\overset{m}{\mathrm{\&Sigma;}}}{({\mathrm{\&mu;}}_i-\mathrm{\&mu;})}^2}---\left(20\right)$

S5.6, low value area judging, the multiple departing from standard deviation by each cell-average difference finds out load-carrying efficiency depression, then have

$Q_i=\frac{{\mathrm{\&mu;}}_i-\mathrm{\&mu;}}{\mathrm{\&delta;}}---\left(21\right)$

S5.7, calculate each town street grid number N _t, with town street for unit, select large principle by GIS software and calculate to belong to area the grid cell number N that each town street comprises _t;

S5.8, in units of the street of town identification effect depression, calculate the mean value that each town street construction land carrying depression unit number departs from whole city's standard deviation multiple;

$\stackrel{\&OverBar;}{Q_t}=\frac{\mathrm{\&Sigma;}{Q}_{\mathrm{it}}}{N_t}---\left(22\right)$

Judge whether the construction land in each town street belongs to the region, depression of load-carrying efficiency thus.