Mobile Phone Data Statistics as Proxy Indicator for Regional

Economic Activity Assessment

Irina Arhipova

1 a

, Gundars Berzins

, Edgars Brekis

, Juris Binde

and Martins Opmanis

Faculty of Business, Management and Economics, University of Latvia, Aspazijas boulevard 5, Riga, Latvia

Latvian Mobile Telephone, Ropazu street 6, Riga, Latvia

The Institute of Mathematics and Computer Science, University of Latvia, Raina boulevard 29, Riga, Latvia

martins.opmanis@lumii.lv

Keywords: Efficiency, Principal Component Analysis, Theil Index.

Abstract: The mobile data analysis is an authoritative source of information for problems solving in the fields of

human activity recognition, population dynamics, tourism, transport planning, traffics measuring, public

administration and other activities and could be the source for valuable information as a proxy indicator.

One of the obstacles to user data from mobile operators is compliance to the General Data Protection

Regulation, so the development of data analytics approach that protects personal data without a necessity to

identify mobility of particular persons was developed, that still provides economically relevant data. In the

present research, the method for the economic activity assessment based on mobile phone data statistics of

any analysed region has been developed. The data of person activity was aggregated at the area of each base

station by the 15 minutes interval, where the activity is defined as the number of outgoing and incoming

calls, sent and received short message service (SMS), as well the number of the unique users. The Latvian

counties have used as a case study where regions were grouped into the similar categories and compared for

two periods: 2015 – 2016 and 2017 by the economic activity efficiency with particular attention to the

seasonality effect for mobile phone activities in counties. It was concluded, that the economic activity of

counties can be estimated and in the particular case positive dynamics of regional development has been

detected.

https://orcid.org/0000-0003-1036-2024

1 INTRODUCTION

Today's working life is intertwined with the use of

mobile phones. They have long ceased to be just an

information exchange tool. Unlike stationary phones

that could be used to get list recipients and analyse

frequency and duration of calls made, in the case of

mobile phones, additional information can also be

obtained about movements of the owner of the

phone over time. Investigating particular mobile

phone activities (the facts themselves, not their

content) can provide insight into the mobility of the

population and its economic activity. The obtained

conclusions can be useful for the decision-making

about regional development and could serve as

metrics characterizing national economy. The

mobile data analysis is an authoritative source of

information for problems solving in the fields of

human activity recognition, population dynamics,

tourism, transport planning, traffics measuring, and

public administration. The authors in previous

research have analysed the mobile positioning data

for residents’ movements (Ahas et al., 2010;

Zonghao et al., 2013), automatic recognition of

population activities (Chetty et al., 2015; Lee and

Cho, 2014), estimation of human trajectories (Hoteit

et al., 2014; Larijani et al., 2015; Liu et al., 2013;

Zilske and Nagel, 2015) and flows (Balzotti et al.,

2018), patterns of population dynamics (Deville et

al., 2014; Trasarti et al. 2015).

The identification of tourists’ destinations

(Alexander et al., 2015; Raun et al., 2016), seasonal

patterns (Ahas et al., 2007; Phithakkitnukoon et al.,

2015), traveler’s preferences (Y. Wang et al., 2018)

and behavior (Z. Wang et al., 2018), evaluation of

tourism sector (Ahas et al., 2008; Kuusik et al.,

Arhipova, I., Berzins, G., Brekis, E., Binde, J. and Opmanis, M.

Mobile Phone Data Statistics as Proxy Indicator for Regional Economic Activity Assessment.

DOI: 10.5220/0007772000270036

In International Conference on Finance, Economics, Management and IT Business (FEMIB 2019), pages 27-36

ISBN: 978-989-758-370-4

2014), travel flow (Ni et al., 2018), tourist

movement patterns (Zhao et al., 2018), trip

modelling (Bwambale et al., 2017), analysis of

number of travelers (Sørensen et al., 2018) and

passengers demand (Hatziioannidu and

Polydoropoulou, 2017) are the popular

investigations problems.

Mobile phone data have used for transport

planning (Elias et al., 2016; Liu et al., 2014), traffic

measurement (Dong et al., 2015; Hongyan and

Fasheng, 2013; Steenbruggen et al., 2016) and

modelling (Oliveira et al., 2017), trajectories

evaluation (Bonnel et al., 2015; Chen et al., 2014)

and predicting travel time (Woodard et al., 2017).

The general problems in urban planning (Ahas and

Mark, 2005; Ricciato et al., 2017; Jonge et al., 2012)

and analysis (Lee et al., 2018), land using (Ríos and

Muñoz, 2017) and smart city development

(Steenbruggen et al., 2015) are solved by mobile

phone data analyzing.

As the basic unit of the mobile network

infrastructure is a cell with its own base station, at

the beginning of each call a specific base station

determines the location of the mobile phone. If,

during a call, the mobile phone moves beyond the

limits of a particular cell, then switching to another

base station takes place. If a person is sufficiently

mobile, then a large number of base stations can be

used to initialize calls. Although more than one base

station may be used during one call, enough

information can be obtained by analyzing only

where the calls have started. If there are several base

stations used for initializing calls during the day,

then disposition of stations quite accurately reveals

the habits of its owner. For example, if

conversations in the mornings and evenings were

initiated from one base station, but in mid-days from

another, it may be assumed that these conversations

geographically outline the person’s working place

and residence.

The data obtained can also be studied on the

basis of information collected by the base stations on

the calls initiated and/or short message service

(SMS) sent. Changes in mobile activities during the

day (week, month) can provide relevant information

about the habits of the region's population. Within

the framework of the study, it was intended to

process data of the mobile telecommunications

operator Latvian Mobile Telephone (LMT) to

identify social factors influencing society and the

national economy. The analysis of the mobile phone

data statistics makes it possible to create different

types of metrics of important processes in the

society. In the original model the study of all mobile

calls, anonymizing only the number was planned.

Therefore, it still would be possible to trace

activities of each particular anonymous person - to

determine which calls were made using the same

number (most likely by the same person), as well as

observe the habits of moving this person.

The General Data Protection Regulation

(GDPR), which is determined by the Law on

Personal Data Processing and started on 25 May

2018, providing uniform rules for the protection of

personal data throughout the European Union (EU).

The Regulation will apply to any company, entity or

organization that processes or stores data from

identifiable individuals living in the EU (European

Parliament, 2016). Taking into account the

requirements of the GDPR within the framework of

this study switching of particular individual calls are

not fixed and tied to the corresponding activity. It is

considered that it is reasonably sufficient to identify

the base stations that ensure connection to the

network. Where with another approach was used to

do not require identification of mobility of a

particular person - the data of person activity was

aggregated at the area of each base station by the 15

minutes interval; where the activity may be outgoing

(calls and sent SMS) as well as incoming (calls and

received SMS). Also, the number of unique users for

each interval is counted. The last one gives insight

about average activity of persons in the region of a

particular base station. Therefore the person data

protection is respected. Undoubtedly, if compared to

the original, a significant part of the data is lost.

However, in the current version, data can still be

used to:

 explore the mobility of people and their

everyday habits (such as leaving home in the

morning and returning to the evenings) as well

as investigate serious demographic processes

(relocation to cities, natural disasters, etc.);

 explore social interactions. Mobile phone

usage models in different layers of society

(gender, age groups, education, etc.) vary

significantly. Knowing behaviour patterns, it

is possible to find out the proportion of

different groups in the region, as well as to

observe changes in the habits (or the

proportion of different groups);

 anticipate economic activity. Mobile service

cost dynamics can quite accurately predict the

upcoming crises.

Call Detail Record (CDR) is a digital data

recorder used for telephone communications or other

equipment for telecommunications, which includes

telephone conversations or other

FEMIB 2019 - International Conference on Finance, Economics, Management and IT Business

telecommunications transactions (for example, text

messages) that are broadcast on the device. CDR

contains call time, duration, call status, caller and

calling subscriber numbers. The CDR does not

contain information about the content of the

conversation. In previous research was concluded

that mobile phone data are suitable and updatable

for the Latvian regional business index development

(Arhipova et al., 2017). The regions with similar

economic activity patterns using mobile

communication data were identified and it was

concluded that mobile phone activities have

statistically significant relationship to regional

economic activity such as Gross Domestic Product

(GDP), number of economically active enterprises

per thousand inhabitants, municipalities budget

expenditures and other (Arhipova et al., 2019). As a

result, the hypothesis that counties and regions with

lower call activity have lower economic activity

compared to other regions with higher call activity

could not be rejected.

The objective of this paper is to develop method

for the economic activity assessment of any region

being researched, based on mobile phone data

statistics. Using previous research results, the first

CDR data analysis was made for the time period

from July 2015 to January 2018, with particular

attention to the seasonality effect for mobile phone

activities in counties. The second, the counties were

grouped into the similar categories based on its

economic activity and compared for two periods:

2015 – 2016 and 2017. The third, the counties’

economic activity efficiency in and its change

dynamics was estimated. The fourth, the region’s

economic activity was estimated by Theil index and

conclusions were made about Latvia region

economic development tendencies.

2 DATA AND METHODS

In addition to the connection attributes already

mentioned, CDR generated automatically by the

mobile network operator initializing mobile phone

call or sending SMS contains information about the

base stations that provides the connection and

calling side. Since the coordinates of the base

stations are known, the location of the persons at the

beginning of the call can be determined with

appropriate precision. Each database entry includes

the following parameters: the total number of calls

and SMS, the total number of unique users, date and

15 minutes time interval in the daytime, antenna

identifier (ID) of the mobile network base station

and its coordinates.

2.1 Data

The database used for the current case study consists

of the roaming data of the mobile phone call

activities of Latvia Mobile Telephone for 30 months

from 25 July 2015 to 20 January 2018, altogether

108 008 160 CDR (Call Data Record) from 1235

base stations for 911 days with 15 minutes time

intervals per hour.

The distribution of network base stations

between Latvia counties and regions was obtained

using its geographical coordinates. An

administrative division includes 110 counties and 9

cities (Riga, Jekabpils, Jelgava, Jurmala, Ventspils,

Liepaja, Daugavpils, Rezekne and Valmiera) in

Latvia, but for statistical and planning purposes six

statistical regions have been formed: Kurzeme,

Latgale, Pieriga, Riga, Vidzeme and Zemgale. Data

analysis shows the difference between the intensity

of call activity on business days and on holidays

(Saturday, Sunday or public holidays), which

characterizes the economic activity of the area

(Arhipova et al., 2017). At the same time, the

maximum of the call activity is observed on noon for

all days (Arhipova et al., 2019), as well the

seasonality effect in summer and winter holidays.

In Figure 1 the total number of calls and SMS for

three cities Jelgava, Jurmala and Ventspils is shown

from August 2015 to December 2017, where in

summer time Jurmala has the highest call activity,

Jelgava has the lowest call activity, but in Ventspils

mobile call activity doesn’t have a strong seasonal

effect. All cities have a seasonal effect in December,

due to the winter holidays (Christmas and New

Year).

Figure 1: The total number of mobile phone activities in

Jelgava, Jurmala and Ventspils.

To analyse the distribution of counties and cities

by economic activity, Principal Component Analysis

(PCA) was used to find out which counties are

Mobile Phone Data Statistics as Proxy Indicator for Regional Economic Activity Assessment

similar to the characteristics of the mobile phone call

activities and group it into the similar categories or

groups. In turn, the Theil index is proposed to

distribute the regions according its economic

activity. It is necessary to stress, that the Theil index

measures the distribution of inequality not only

within the same group, but also between different

groups (Bellù and Liberati, 2006), for example,

regional inequalities.

2.2 Grouping of Counties by Similar

Economic Activity

Before grouping the counties into the similar

categories based on its economic activity, the 119

variables as the linear combination of the total

number of mobile phone activities and the total

number of unique users for all counties were

developed depending on day during the 2015 – 2017

time periods. The PCA was applied using Varimax

rotation separately for two time periods:

 from 25 July 2015 to 20 January 2017 and

 from 21 January 2017 to 20 January 2018.

In the first time period 67.6% of the total

variance is described by the first two principal

components (PC), where the 1

PC has high values

in business days as the counties with higher

economic activity, but the 2

PC has high values

holidays as the counties with lower economic

activity (Arhipova et al., 2019). In the next time

period from 21 January 2017 to 20 January 2018 the

results of applied PCA shows that 71.0% of the total

variance is described by the first two principal

components. To find out the interpretation of

principal components, their average values were

calculated based on weekdays (Fig.2) and months

(Fig. 3).

Figure 2: Principal components average values depending

on weekdays in 2017.

It can be concluded that the 1

PC has the

highest values on business days and lower values on

holidays and summer months. In contrast, the 2

has lower values on business days and higher values

on holidays and summer months. The component

loadings, which are the correlations of the observed

119 variables with the first two principal

components, are used to interpret the meaning of

components. It is hypothesized that counties and

cities with higher economic activity correlate highly

with the 1

PC, but with lower economic activity

correlate highly with the 2

PC. The distribution of

the counties and cities by economic activity can be

shown using the loading plot an orthogonal solution.

Latvian counties and cities are grouped into 8 groups

according to their economic activity. These groups

allow you to understand the profile of each county

and depend on the economic activity on business

days and on holidays (Saturday, Sunday or public

holidays), as well as seasonality effect.

Figure 3: Principal components average values depending

on months in 2017.

The summary of the proposed counties’ groups,

using the two components loadings or correlation

coefficients, is shown in Table 1.

Table 1: Summary of counties’ groups.

Group

Economic activity on

business days

PC value)

holidays

PC value)

Hard Workers

high

(0.8 – 1.0)

average low

(0.0 – 0.2)

Congruent

high

(0.8 – 1.0)

average

(0.2 – 0.6)

Moderate

average

(0.4 – 0.8)

average

(0.2 – 0.6)

Disinterested

low

(0.0 – 0.4)

average

(0.2 – 0.6)

Holidaymakers

average low

(0.4 – 0.6)

average high

(0.6 – 0.8)

Party Makers

low

(0.0 – 0.4)

high

(0.6 – 1.0)

Hedonists

lowest

(-0.2 – 0.0)

highest

(0.8 – 1.0)

Phenomenon

average

(0.4 – 0.8)

average low

(0.0 – 0.2)

FEMIB 2019 - International Conference on Finance, Economics, Management and IT Business

In order to evaluate the effectiveness of the

economic activity strategy chosen by each county,

40% to 100% efficiency curves were constructed. It

is necessary to stress, that the amount of variance in

each variable explained by the principal components

or the component communalities are computed by

taking the sum of the squared loadings for that

variable, where component loadings are the

correlations of the observed variables with the

principal component.

Therefore the efficiency criterion EC was

calculated according to the formula (1):

 















(1)

where 





is correlation coefficient of the observed

variable (linear combination of the total number

of mobile phone activities and the total number of

unique users in n

county) with the first principal

component loadings, but 





is correlation

coefficient of the observed n

variable with the

second principal component loadings.

2.3 Theil Index for Regions Economic

Activity Evaluation

Theil index is a measure of overall inequality and

primarily used to measure economic inequality.

According the proposed hypothesis that mobile

phone activities have statistically significant

relationship to region economic activity, Theil index

is calculated for the each county mobile phone

activities data, aggregated by day for the time period

from 25 July 2015 to 20 January 2017, using the

formula (2):





































(2)

where y

is total number of mobile phone activities

on a particular day, n is the number of days, and  is

the average number of mobile phone activities for

observed time period. The higher is the value of the

Theil index, the greater is the data inequality.

Theil index is used to compare changes in

mobile phone activities on business days and

holidays. The higher the value of the Theil index, the

higher the inequality between the number of mobile

phone activities on business days and holidays.

It should be noted that Theil indexes are equal

for the two possible cases: high mobile phone

activities on business days and low on holidays or

low mobile phone activities on business days and

high on holidays.

Theil index was calculated for six statistical

regions in Latvia: Kurzeme, Latgale, Pieriga, Riga,

Vidzeme and Zemgale, using the number of mobile

phone activities that are grouped by days for each

statistical region.

3 RESULTS

Based on the PCA obtained results, Latvian 110

counties and 9 cities are grouped into 8 groups

(Table 1) according to their mobile phone activity

for two periods: 2015 – 2016 and 2017.

The efficiency of the economic activity strategy

was calculated, using formula (1), and compared by

40% to 100% efficiency curves. The mobile phone

activity of six regions was estimated by Theil index

and the obtained results have compared with

counties’ mobile phone activity for the next Latvia

region economic development tendencies evaluation.

3.1 Latvian Counties Distributions by

Groups

Latvian 110 counties and 9 cities distribution by

groups, according to their mobile phone activity for

time period from 25 July, 2015 to 20 January, 2017

is shown on Figure 4.

The first groups “Hard Workers” is characterized

by high activity on business days, but on average

low activity on holidays. It is the driving force

behind the Latvian economy, but does not fully

exploit the holiday potential. The counties are highly

dependent on fluctuations in economic activity, and

it is necessary to develop the service sector. For

example, the capital of Latvia the city of Riga is the

central metropolis of the Baltic States, the

international level port and infrastructure hub (The

Freeport of Riga).

Riga is characterized by transit and logistics

companies, developing social infrastructure. At the

same time, Riga is a monocentric city, which

insufficient uses tourism potential, only 5% of small

and medium enterprises (SMEs). Riga has a

congested traffic infrastructure and a population

reduction. The city of Jelgava is only 40 km far from

to Riga and has excellent infrastructure, besides

there is situated the Latvia University of Life

Sciences and Technologies. However, there is small

business activity on holidays and insufficient leisure

opportunity, only 4% of SMEs work in tourism. The

second group “Congruent” of Latvia's counties is

Mobile Phone Data Statistics as Proxy Indicator for Regional Economic Activity Assessment

Figure 4: Latvian 110 counties and 9 cities distribution by groups in 2015 – 2016.

characterized by high and moderate activity on

business days and average activity on holidays. The

group is characterized by balanced development, but

insufficient resources for the next breakthrough.

Depending on the priorities, it is necessary to

develop the production or service sector, but the

wrongly selected priorities can be broken down by

the available resources. For example, the city of

Ventspils has a developed production and logistics

sector, a well-developed port, Ventspils University

College and in tourism sector there is 5% of SMEs.

The group “Holidaymakers” is characterized by

an average low activity on business days, but on

average high activity on holidays. The counties

sufficiently use the holiday potential, but

insufficiently on business days. It is necessary to

develop the production sector and change the

county's development strategy. For example, the city

of Jurmala is a historic resort town near Riga with

favourable strategic positioning, where 7% of SMEs

are in the tourism sector. However, it does not fully

use its potential on holidays.

In both cities of Riga and Jelgava from group

“Hard Workers” exist high activity on business days,

but on average low activity on holidays, as well a

strong negative seasonal effect in summer time.

The city of Ventspils from the second group

“Congruent” is characterized by high and moderate

activity on business days and average activity on

holidays and summer time. In turn the city of

Jurmala from the group “Holidaymakers” has an

average low activity on business days, but on

average high activity on holidays, as well the strong

positive seasonal effect in summer time.

The average number of mobile phone call

activities of LMT per month in Riga, Jelgava,

Ventspils, and Jurmala in 2017 is given in Figure 5.

Figure 5: The average number of mobile phone call

activities in cities per month.

The group “Moderate” is characterized by the

average economic activity on all days. It is

characterized by uniform activity, where the

resource potential is not sufficiently used. It is

necessary to increase labour productivity and

economic potential, otherwise economic activity and

regional development will decrease.

The group “Disinterested” is characterized low

activity on business days and average activity on

FEMIB 2019 - International Conference on Finance, Economics, Management and IT Business

holidays. There is a holiday potential, but low

economic activity on business days. It is necessary

to develop the service sector and to change the

development strategy of the region, threatening the

region's degradation.

The group “Party Makers” is characterized by

low activity on business days, but high activity on

holidays. It is necessary to develop the production

sector and change the region's development strategy.

There is a high dependence on the purchasing power

of the population.

The group “Phenomenon” is characterized by

average activity on business days and moderate low

activity on holidays, where in 2015 – 2016 it was

Vilanu county, but in 2017 - Rugaju county.

The “Hedonists” group is characterized by the

lowest activity on business days, but the highest

activity on holidays. There is no economic potential

for the manufacturing sector. It is necessary to

develop the production sector and change the

region's development strategy. There is a maximum

dependence on the purchasing power of the

population. For example, Rucavas county has a

strong positive seasonal effect in summer months

and it is unique because the number of mobile phone

activities on holidays is higher than in business days

(Fig. 6).

Figure 6: The average number of mobile phone call

activities in Rucavas county per month and week days.

3.2 Latvian Counties Economic

Activity Efficiency

The distribution of the counties by economic activity

can be shown using the principal components loads

with data aggregated by days. Latvian counties are

divided into 8 groups according to their economic

activity.

The negative 2

PC values in group “Hedonists”

means, that the mobile phone activity during the

holiday in absolute values is higher, that in business

days. It is a character only for counties from group

“Hedonists”: Saulkrastu county in 2015 – 2016 and

Rucavas county in 2015 – 2016 and 2017 time

periods.

Groups allow you to understand the profile of

each county, but the efficiency curve makes it

possible to assess the effectiveness of the strategy

chosen by each county (Fig. 7).

PC: high values in business days

2015 – 2016 2017

PC: high values in holidays

Figure 7: Distribution of Latvian counties in groups after the effectiveness of economic activity.

Mobile Phone Data Statistics as Proxy Indicator for Regional Economic Activity Assessment

In 2017, in comparison with 2015 - 2016, the

economic activity of Latvian counties is improving,

as indicated by the distribution of groups, for

example, the number of counties included in the

group “Disinterested” decreased two times.

The distribution of 119 Latvian counties in 2017

according to the efficiency criterion EC displays that

the number of the counties lies between the two

different efficiency curves are the following:

 16 counties from 95% to 100%;

 31 counties from 90% to 95%;

 36 counties from 80% to 90%;

 36 counties from 45% to 80%.

In comparison in 2015 – 2016 time period there

were two counties with efficiency less than 45%.

3.3 Distribution of Regions by

Economic Activity using Theil

Index

In order to evaluate the regional differences in call

activity between business days and holidays, Theil

index E(1) for each region was calculated for period

from 1 August 2015 to 31 December 2016.

The highest index value indicates a greater

difference in call activity between business days and

holidays, which indicates a higher economic

activity. According to the Theil index the higher

economic activity is in Riga, than Kurzeme and

Pieriga, but Latgale and Vidzeme have the lowest

results in economic activity (Fig. 8).

Figure 8: Theil index for Latvian regions in 2015-2016.

The similar results were obtained, using the

PCA, which shows that cities and regions with lower

call activity have lower economic activity compared

to other regions with higher call activity (Arhipova

et al., 2019).

To estimate the dynamics of the difference

between regional call activities in business days and

holidays, the Theil index for each region is

calculated per month. There is a close correlation

between regions’ Theil index change dynamics by

months, except city of Riga, and indicates the related

tendency of regions economic development.

4 CONCLUSIONS

The number of mobile phone call activities is an

indicator of economic activity for counties and

regions and could be used to develop a reliable tool

for continuous and dynamic monitoring of the

region’s economic performance.

There is a seasonal effect in mobile phone

activities, as well a significant difference between

business days and holidays, that indicates and

distinguishes counties and regions with different

economic activity patterns.

Data obtained from Latvian counties resulted in

the identification of distinct 8 groups representing

the unique pattern of economic activity. The

efficiency curve makes it possible to evaluate the

effectiveness of the strategy chosen by each county.

Characteristics of the distribution of Latvian

regional groups depend on economic activity on

business days and holidays, as well on seasonal

effect.

In 2017, in comparison with 2015 - 2016, the

economic activity of Latvian counties is improving,

as indicated by the distribution in groups, for

example, the number of counties included in the

group "Disinterested" decreased by two times.

The Theil index characterizes the difference in

mobile phone activities between business days and

holidays. The results of PCA and Theil Index give

the same distribution of counties and regions by its

economic activity.

The authors propose the regional development

index as a real-time or periodic monitoring tool.

Developed method provides a practical tool for

regional governments in keeping track on strategy

implementation and strategic gap analysis.

This method also provides dynamic visualization

of strategic direction particular municipality have

achieved between periods of measurement and can

be used as central additional performance indicator

regional governments measure regularly.

The developed method tested on Latvian mobile

telephone data sets as proxies, and the regional

development index was created.

Regional economic activity efficiency evaluation

is based on mobile phone data statistics, taking into

account the requirements of the GDPR, where

switching of particular individual calls are not fixed,

and the personal data protection is respected.

FEMIB 2019 - International Conference on Finance, Economics, Management and IT Business

ACKNOWLEDGEMENTS

This work was supported by the University of Latvia

and LMT Ltd. [grant number AAP2016/B089].

REFERENCES

Ahas, R., Aasa, A., Silm, S., Tiru, M., 2010. Daily

rhythms of suburban commuters’ movements in the

Tallinn metropolitan area: Case study with mobile

positioning data. Transportation Research Part C,

18(1), pp.45–54.

Ahas, R., Aasa, A., Roose, A., Mark, U., Silm, S., 2008.

Evaluating passive mobile positioning data for tourism

surveys: An Estonian case study. Tourism

Management, 29, pp.469–486.

Ahas, R., Aasa, A., Mark, U., Pae, T., Kull, A., 2007.

Seasonal tourism spaces in Estonia: Case study with

mobile positioning data. Tourism Management, 28,

pp.898–910.

Ahas, R., Mark, U., 2005. Location based services — new

challenges for planning and public administration?

Futures, 37, pp.547–561.

Alexander, L., Jiang, S., Murga, M., Gonzalez, M. C.,

2015. Origin – destination trips by purpose and time of

day inferred from mobile phone data. Transportation

Research Part C, 58, pp.240–250.

Arhipova, I., Berzins, G., Brekis, E., Opmanis, M., Binde,

J., Steinbuka, I., Kravcova, J., 2019. Pattern

Identification by Factor Analysis for Regions with

Similar Economic Activity Based on Mobile

Communication Data. Advances in Intelligent Systems

and Computing, 886, pp.561–569.

Arhipova, I., Berzins, G. Brekis, E., Kravcova, J., Binde,

J., 2017. The methodology of region economic

development evaluation using mobile positioning data.

In D. Cingula, ed. 20th International Scientific

Conference on Economic and Social Development.

Prague: Varazdin Development and Entrepreneurship

Agency, Varazdin, Croatia. City of Prague. University

North, Koprivnica, Croatia. Faculty of Management

University of Warsaw, Warsaw, Poland, pp. 111–120.

Balzotti, C., Bragagnini, A., Briani, M., Cristiani, E.,

2018. Understanding Human Mobility Flows from

Aggregated Mobile Phone Data. IFAC-PapersOnLine,

51(9), pp.25–30.

Bellù, L.G., Liberati, P., 2006. Describing Income

Inequality. Theil Index and Entropy Class Indexes,

Available at: www.fao.org/tc/easypol.

Bonnel, P., Hombourger, E., Olteanu-Raimond, A.-M.,

Smoreda, Z., 2015. Passive mobile phone dataset to

construct origin-destination matrix: potentials and

limitations. Transportation Research Procedia, 11,

pp.381–398.

Bwambale, A., Choudhury, C.F., Hess, S., 2017.

Modelling trip generation using mobile phone data: A

latent demographics approach. Journal of Transport

Geography, (August), p.article in press.

Chen, C., Bian, L., Ma, J., 2014. From traces to

trajectories: How well can we guess activity locations

from mobile phone traces? Transportation Research

Part C, 46, pp.326–337.

Chetty, G., White, M., Akther, F., 2015. Smart Phone

Based Data Mining For Human Activity Recognition.

Procedia Computer Science, 46, pp.1181–1187.

Deville, P., Linard, C., Martin, S., Gilbert, M., Stevens,

F.R., Gaughan, A.E., Blondel, V.D., Tatem, A.J.,

2014. Dynamic population mapping using mobile

phone data. PNAS (Proceedings of the National

Academy of Sciences of the United States of America),

111(45), pp.15888–15893.

Dong, H., Wu, M., Ding, X., Chu, L., Jia, L., Qin, Y.,

Zhou, X., 2015. Traffic zone division based on big

data from mobile phone base stations. Transportation

Research Part C, 58, pp.278–291.

Elias, D., Nadler, F., Stehno, J., Krosche, J., Lindorfer, M.,

2016. SOMOBIL – improving public transport

planning through mobile phone data analysis.

Transportation Research Procedia, 14, pp.4478–4485.

European Parliament, 2016. REGULATION (EU)

2016/679 OF THE EUROPEAN PARLIAMENT

AND OF THE COUNCIL of 27 April 2016 on the

protection of natural persons with regard to the

processing of personal data and on the free movement

of such data, and repealing Directive 95/46/EC

(General Da. Official Journal of the European Union,

L 119, pp.1–88.

Hatziioannidu, F., Polydoropoulou, A., 2017. Passenger

Demand And Patterns Of Tourists’ Mobility In The

Aegean Archipelago With Combined Use Of Big

Datasets from Mobile Phones and Statistical Data

from Ports and Airports. Transportation Research

Procedia, 25, pp.2309–2329.

Hongyan, G., Fasheng, L., 2013. Estimating freeway

traffic measures from mobile phone location data.

European Journal of Operational Research, 229(1),

pp.252–260.

Hoteit, S., Secci, S., Sobolevsky, S., Ratti, C., Pujolle, G.,

2014. Estimating human trajectories and hotspots

through mobile phone data. Computer Networks, 64,

pp.296–307.

Jonge, E. de, Pelt, M. van, Roos, M., 2012. Time patterns,

geospatial clustering and mobility statistics based on

mobile phone network data. The Hague: Statistics

Netherlands.

Kuusik, A., Nilbe, K., Mehine, T., Ahas, R., 2014.

Country as a free sample: the ability of tourism events

to generate repeat visits. Case study with mobile

positioning data in Estonia. Procedia - Social and

Behavioral Sciences, 148, pp.262–270.

Larijani, A.N., Olteanu-Raimond, A.-M., Perret, J., Bredif,

M., Ziemlicki, C., 2015. Investigating the mobile

phone data to estimate the origin destination flow and

analysis; case study: Paris region. Transportation

Research Procedia, 6, pp.64–78.

Lee, K.-S., You, S.Y., Eom, J.K., Song, J., Min, J.H.,

2018. Urban spatiotemporal analysis using mobile

phone data: Case study of medium- and large-sized

Mobile Phone Data Statistics as Proxy Indicator for Regional Economic Activity Assessment

Korean cities. Habitat International, 73 (October

2017), pp.6–15.

Lee, Y.-S., Cho, S.-B., 2014. Activity recognition with

android phone using mixture-of-experts co-trained

with labeled and unlabeled data. Neurocomputing,

126, pp.106–115.

Liu, F., Janssens, D., Wets, G., Cools, M., 2013.

Annotating mobile phone location data with activity

purposes using machine learning algorithms. Expert

Systems With Applications, 40(8), pp.3299–3311.

Liu, F., Janssens, D., Cui, J.X., Wang, Y.P., Wets, G.,

Cools, M., 2014. Building a validation measure for

activity-based transportation models based on mobile

phone data. Expert Systems with Applications, 41(14),

pp.6174–6189.

Ni, L., Wang, X.C., Chen, X.M., 2018. A spatial

econometric model for travel fl ow analysis and real-

world applications with massive mobile phone data.

Transportation Research Part C, 86(April 2017),

pp.510–526.

Oliveira, E.M.R., Viana, A.C., Naveen, K.P., Sarraute, C.,

2017. Mobile data traffic modeling: Revealing

temporal facets. Computer Networks, 112, pp.176–

193.

Phithakkitnukoon, S., Horanont, T., Witayangkurn, A.,

Siri, R., Sekimoto, Y., Shibasaki, R., 2015.

Understanding tourist behavior using large-scale

mobile sensing approach: A case study of mobile

phone users in Japan. Pervasive and Mobile

Computing, 18, pp.18–39.

Raun, J., Ahas, R., Tiru, M., 2016. Measuring tourism

destinations using mobile tracking data. Tourism

Management, 57, pp.202–212.

Ricciato, F., Widhalm, P., Pantisano, F., Craglia, M.,

2017. Beyond the “single-operator, CDR-only”

paradigm: An interoperable framework for mobile

phone network data analyses and population density

estimation. Pervasive and Mobile Computing, 35,

pp.65–82.

Ríos, S.A., Muñoz, R., 2017. Land Use detection with cell

phone data using topic models: Case Santiago, Chile.

Computers, Environment and Urban Systems, 61,

pp.39–48.

Sørensen, A.Ø., Bjelland, J., Bull-Berg, H., Landmark,

A.D., Akhtar, M.M., Olsson, N.O.E., 2018. Use of

mobile phone data for analysis of number of train

travellers. Journal of Rail Transport Planning &

Management, 8(2), pp.123–144.

Steenbruggen, J., Tranos, E., Nijkamp, P., 2015. Data

from mobile phone operators : A tool for smarter

cities ? Telecommunications Policy, 39(3-4), pp.335–

346.

Steenbruggen, J., Tranos, E., Rietveld, P., 2016. Traffic

incidents in motorways : An empirical proposal for

incident detection using data from mobile phone

operators. Journal of Transport Geography, 54,

pp.81–90.

Trasarti, R., Olteanu-Raimond, A.-M., Nanni, M.,

Couronne, T., Furletti, B., Giannotti, F., Smoreda, Z.,

Ziemlicki, C., 2015. Discovering urban and country

dynamics from mobile phone data with spatial

correlation patterns. Telecommunications Policy, 39(3-

4), pp.347–362.

Wang, Y., Correia, G.H. de A., Arem, van B.,

Timmermans, H.J.P., 2018. Understanding travellers’

preferences for different types of trip destination based

on mobile internet usage data. Transportation

Research Part C, 90(October 2017), pp.247–259.

Wang, Z., He, S.Y., Leung, Y., 2018. Applying mobile

phone data to travel behaviour research: A literature

review. Travel Behaviour and Society, 11, pp.141–

155.

Woodard, D., Nogina, G., Koch, P., Racz, D., Goldszmidt,

M., Horvitz, E., 2017. Predicting travel time reliability

using mobile phone GPS data. Transportation

Research Part C, 75, pp.30–44.

Zhao, X., Lu, X., Liu, Y., Lin, J., An, J., 2018. Tourist

movement patterns understanding from the perspective

of travel party size using mobile tracking data: A case

study of Xi’an , China. Tourism Management,

69(February), pp.368–383.

Zilske, M., Nagel, K., 2015. A simulation-based approach

for constructing all-day travel chains from mobile

phone data. Procedia Computer Science, 52, pp.468–

475.

Zonghao, R., Dongyuan, Y., Zhengyu, D., 2013. Resident

Mobility Analysis Based on Mobile-Phone Billing

Data. Procedia - Social and Behavioral Sciences,

96(Cictp), pp.2032–2041.

FEMIB 2019 - International Conference on Finance, Economics, Management and IT Business