Various complicated events in nature are not simple to be formed and also reproduced for specific events are not always constant, expected and also linear but effective, non-linear and complicated. Using the relatively suitable description or image to explore the effective method and also pattern is important for us to examine its working device. For instance, city development is the complicated interactive method but it can be essentially determined by internal or exterior circumstances which are deeply compared to a city. Zipfs law acknowledged as a standard rank-size relationship, is one of proper standards used for describing complicated events. It was originally recommended by German physicist Felix Auerbach in 1913, and also further defined by the American linguist George Kingsley Zipf in 1949. Zipfs law is usually indicated by y=??1 where “y” is city size and “x” describes city rank. If city size is arranged from highest to least according to their population, it can be recognised that the first biggest city is usually twice as big as the second biggest city, three times as much as third one and so on. As searching the completeness of Zipfs law, to some degree, can help us to assemble many parts of data used for profoundly investigating Zipfs law pattern and its underlying mechanism, therefore the experimental investigation could be significant and important. That investigation is going to prove whether Zipfs law can be true for the natural cities selected from location-based social media data. Section one supplies a short background to the knowledge of Zipfs law and why considering Zipfs law is important. Section displays the Heavy tail distribution, Power law and Zipfs law in this study and so one.
There must be two basic issues circling the investigation on Zipfs law in terms of city investigations. The first issue relates to whether Zipfs law can be maintained for several countries or regions, which is connected with whether Zipfs law is general. The second issue about Zipfs law is why it should exist and why the hierarchical system is so popular, which more leads to the device of Zipfs law and its underlying meaning. People regularly arrange involved when there appears to be an related or natural event happening both in nature and society but including much various forms. For instance, various events in both natural and also unnatural ways can display related Zipfs law patterns and described systems but they are extremely unlike in terms of their element units and communication factors. C?rdoba in 2008 and also Krugman in 1996 showed out that recognising Zipfs law patterns could be very important in further demonstrating why such hierarchical events seem. Virkar and also Clauset inj 2014 recognised this duration and additional registered the power-law-like events, due to scale free assets, can be considered as one of the significant pieces of evidence of investigating and resolving that underlying and unusual development of methods.
Nevertheless, it is a continuous method. The first action for utmost empirical investigations in terms of Zipfs law was to investigate whether Zipfs law is remarkably comprehensive, which additional leads to whether Zipfs law is important to be modelled as an empirical description. Profoundly speaking, investigating the universality of Zipfs law is really a probative method, by which several hypothesizes could be shown as interactive representatives or information. Proving the universality of Zipfs law, in reality, can help us to investigate commonalities of Zipfs law behaviour. By investigating the commonalities, it is possible to determine underlying systems and tools. For instance, when Zipfs law can not be recognised in several systems, it means that the system could have several outlying or internal circumstances that can more or less influence emergence of Zipfs law. Therefore, investigating the outlying and internal factors maybe help us to describe why Zipfs law was not in working in such case and why Zipfs law can be real for others. The pieces of evidence and the empirical method can become an essential concluding to improve the knowledge of Zipfs law.
In names of validation of Zipfs law, there should be various investigations on this topic in terms of city view. For instance, Ioannides and also Overman in 2003 used data for metro in the United States to examine the validity of Zipfs law; Soo in 2005 examined Zipfs law using new data on 73 countries and two calculating systems; C?rdoba in 2008 observed various limitations for civil parameters and then introduced a regular urban model that can be well described by Zipfs laws; Peng in 2010 investigated the validity of Zipfs law in data set of Chinese city sizes from 1994 through 2004 handling rolling sample regression techniques; Jiang and also Jia in 2011 tested the validity of the Zipfs law for natural cities applying street connections and blocks;
Jiang, Yin and also Liu in 2014 discovered out that both city sizes and city numbers obtained from nightlight description can especially exist the Zipfs law.Between the preceding investigations about the validation of Zipfs law, two papers – Jiang and Jia in 2011; Jiang, Yin and also Liu in 2014 are very unique, for city sizes analysed from these two papers were not removed with regular census-imposed data and official data.
On the opposite, the two documents built a new kind of city, specifically, natural cities working extensive bottom-up data and head or tail breaks selection law. Acknowledging that the created outlines of natural cities were not as perfect as original city limits, it is very important to consider why such natural cities can yet fit Zipfs law. Does that mean that natural cities are a good example for us to investigate Zipfs law? Can Zipfs law yet be valid for natural cities obtained from other data? In the thesis, different natural city guides are created and the universality of Zipfs law is considered correspondingly.
The 21st age is the age of big data. Huge big data collected by remotely sensed data, Global Position System swimming data, and Volunteer Geographic Information data obtained from using Google maps and OpenStreetmap, can present several opportunities in academic investigations and sensible purposes. Particularly, the 21st age has observed that the growth of high tech and the discovery of internet assistance has extremely transformed humans everyday life and way of study. Location based social media assistance helped from smartphones andalso the World Wide Web is displaying frequently widespread. The emerging of location based social media systems, such as Facebook and Twitter, allow people to comfortably share their data on a website and also search where they are or who have been nearby. From the investigation point of view, location based social media assistance can give some excellent insight to achieve human actions and agreement by placing society oriented networks. For instance, Cranshaw in 2012 proved a clustering model and methodology for investigating the city patterns applying social media data.
To completely investigate the development of cities and its underlying rule and influences, Jiang andd Miao in 2015 suggested the new kind of city description, specifically, natural cities using location based social media data. Unlike earlier city definition that cities were determined applying statistics required data and subjective methods, the natural cities were described by the series of spatially collected geographic issues and were named by head or tail breaks division rule, which are an more natural and objective method. Based on the emerging natural cities, the central evidence of the thesis aims to verify the whole of Zipfs law for the social spatial natural cities selected from location based social media data. In order to better recognise Zipfs law from various aspects, city numbers, city sizes and population are considered. Some specific questions in terms of location based social media data are also put forward. For instance, can the characteristics of Zipfs law change over the spatial temporal scales? Can Zipfs law yet be valid for all social media data? What are the variations among location based data and early nightlight description Jiang, Yin and Liu in 2014 concerning validation of Zipfs law?
Now we are starting a make new big data age full of possibilities and also challenges. There are a large number of developments that have become the place in data collecting and data interpretation. Big data is not only of big size as the name symbolises, but it is also very diverse in terms of its data source, data models and entities described. The development of big data has dramatically increased the conventional way of thinking from data which can be readily obtained, collected and analyzed onto huge data building difficult to store analyze and imagine. The use of big data is important as it has been discovered of vital significance in various informative fields and methods such as physiography, sociology, economics, mathematics, physics.
In particular, it is very hard for big data to be determined correctly. That obtaining said, big data is usually labelled as a complicated and also flexible data, which are large in amount; high in speed; diverse in variety and exhaustive in field. In current years, many descriptions of the big data have been introduced one after extra. For instance, Hashem in 2015 estimated that big data is a collection of systems which need new forms and ways to achieve. To correctly display the real parts of the big data, Zikopoulos in 2013 further suggested a relatively expansive description of the big data: Volume, Varity and Velocity -Figure 1.