 Then, now I have the pleasure to announce our friend Alexander to be the moderator. So I hope for him the same success as to our friend Zava Zava this morning. Thank you very much. Thank you Mr. Chair. Very good afternoon to all of you. I'm really very happy to be here with you today and for me it's a pleasure to moderate this session on big data for measuring the formation society. Before I give the floor and announce the speakers for this panel, I would like to recall that last week we had the fourth UN Big Data Conference for official statistics that took place in Bogota in Colombia and I would like to just mention that the main topic of this conference was related to the innovation and modernization of the national statistical systems through trusted data collaboratives and in this context the use of big data to complement traditional data sources for the production of official statistics is key to national statistical offices to take full advantage of all this massive data that we have in these big data sources and as you know these big data sources comes from the ICT industry, the telecom and internet providers, social media companies, Twitter, Facebook for instance, mobile and sensor data tracking data, of course private sector transaction data so we have a lot of data sources that we could take benefit of producing official statistics if we have shared data policies if we have in place legal frameworks for data sharing data protection of course data privacy so all these issues I think that it is an opportunity to for us to discuss in this session and I think that it also important to mention that national statistical offices will be under increasing pressure to produce high quality in a timely manner on a wide range of areas often with limited resources so innovation and modernization of national statistical office is of course of utmost importance and data sharing and the use of big data source is really key in this context well this session will showcase the results of the ITU pilot project and big data for measuring the information society and in particular we are going to address some key issues that I hope that our panelists will address and discuss the results of this project and I believe that the ITU pilot project definitely generated a lot of insights from using big data sources to produce ICT related indicators vis a vis the traditional data collection methods that we have in our national statistical systems also I think that we are going to see insights from this project that can be turned into actionable knowledge and support efforts to bridge the digital divide also we are going to see the challenge these pilot countries faced in using those big data sources in terms of negotiating with the data providers and also on how to overcome these challenges and also on how to benefit from the use of big data for producing official and also complement official ICT related statistics so without a father a Jew may I invite Mr. Louis G. Cezari who is the head of the big date implementation at Vodafone Group to give his address Mr. G. Cezari you have 15 minutes you have the floor welcome to everyone here after lunch so Facebook reminded me that two years ago today I was packing up my apartment in New York City getting ready to move to London to embark on this big data journey at Vodafone at the time I had just finished building a big data capability from the ground up at a large U.S.-based insurance firm and everyone around me was asking me are you crazy you just finished building a big data capability why are you going back to the beginning and building it again and the answer is that we as telco providers literally connect the world and we have a tremendous opportunity more than anyone else I believe to shape the future and to really improve the customer experience through the use of big data analytics and the slides are not showing in the back can you help well we can continue with or without slides hopefully at some point you'll see the slides where are we on our journey through big data at Vodafone right now we are in nine countries in the world to be 15 by the end of the year and in all Vodafone markets by the end of next year and we have put the customer experience first and foremost in our big data initiatives focus it's not only about what we do but also about how we do it and we build privacy by design into everything that we do with big data our privacy officers are literally sitting at the table with us day by day involved in the process from beginning to end we also pride ourselves on being just as innovative in our technology security as we are with our data science algorithms it is a point of competitive differentiation in this space and it is a responsibility to our customers and while we believe that everything we do is for the best interests of our customers it is ultimately their choice and we give them that choice through informed consent are you coming for tech support and miraculously we have slides thank you so what are we focusing on at Vodafone at Vodafone we're focusing on helping people connect and communicate through big data both in traditional ways and in new ways we spend the bulk of our time on strengthening the customer experience and delivering what we call the three P's of the digital customer experience predictive proactive and personalized service to all of our 450 million customers throughout the world we are also using big data to optimize the machine that is a telco networks technology other areas of operation but we're not stopping there we're also using big data to create new insights and solutions you may have seen last week the launch of our internet of things suite of products the V suite those needs were directly uncovered through big data analytics and big data analytics is helping us to solve those needs for our customers with needs for our giga cube portable internet our pet tracker our general tracker and then our final area of focus is innovation and social good and we have a director of research but we don't view this as one person's job we view this as every person's job and all of our data scientists and every member of our team can spend 15 percent of their time working on initiatives that are important to them and have social benefit so how do we make all of this happen having built big data capabilities globally now in two countries I look at four essential pillars of how to set up a big data capability the first and most important of those are people while vendors always will have a role if you want to have competitive advantage in big data you must own the capability in-house we have data scientists and data engineers in every Vodafone country supported by a small central team ag group we promote an innovative culture and cross market tribes we view data as a competitive advantage and like everyone else we're looking to have the widest possible internal and external data sources where we really place emphasis though is how do we make data usable and how do we make it automated telco data more than any other data I've seen is extremely messy extremely large it's about boiling the data down to the usable bits and automating the process so that we have this available to us all the time and for all use cases in terms of process we are subscribed to agile methodology we work in two weeks sprints and we have an open code repository even where we cannot share data across countries we can share code across countries so that a data scientist in Turkey or in Tanzania can take something that was developed in Germany and modify the code to fit his or her particular local market needs another change that we firmly believe in is treating customer permissions as an asset too often we view this as just a legal function when in reality it should be a marketing function we should be giving customers clear choice and clear reason to say yes please use my data to improve my experience finally last but not least in the the four essential ingredients is technology I'm not going to give you a list of recommended technology tools because they will probably all change in three to six months from now the key for us is having a coordinated architectural approach across Vodafone markets and the flexibility to continuously upgrade our tools as needed with a quick experimental approach so what have we been focusing on in Vodafone in the past a few years what are some of our initial use cases we placed greatest emphasis on customer retention you know big data is all about the customer experience and we started with a group of customers who had the most need those who were at risk of leaving those who had a poor experience and we use big data to come up with personalized offers that would address those customers pain points and get them to stay and I'm happy to say we have big data powered retention models in all of our markets and that the best performing of them can convince two out of three detractors to remain with Vodafone and become at least passives and hopefully promoters our convergence has another big area of focus for us through the big data initiative for the first time we were able to break down silos from a customer perspective there is nothing more frustrating than being treated like your four different people if you have one mobile account another mobile account a fixed account and a tv account with the same provider now that we have all of that data in one place we've moved from a single customer view to a household view and we're addressing our customers needs and we're coming up with better ideas around new product development as well a third area that was a quick hit for us was fraud detection through big data approaches we can detect fraud nine times faster than we used to and we can stop it and we can reinvest the savings into improving the customer experience through value add but we're not only doing new analytical efforts we are delivering them to customers in new ways as well all of our work is or will be integrated with the my Vodafone app with chat bots and with robots we are using big data to power what will be a new digital experience for our customers I also want to talk very briefly about the work we're doing for social good because we can really use big data to change the world and social good is something our customers want us to do you can see some statistics on the screen and this is from a report published by the Vodafone Foundation in 2015 you can download it online and among Europeans you have more than half of people who would like us to use their data to help them and others to do things that would improve the environment and societies and one of the main ways we've been doing that is through EMPESA which is Vodafone's project product that is live in nine countries with 32 million active customers this is a money transfer initiative that is the ultimate example of doing well by doing good and in the process has lifted 2% of households in Kenya out of poverty we were named number one in fortune magazines change the world list in 2015 and we are committed to continuing to do these types of projects through the Vodafone Foundation and through other partnerships so in closing I would like to say that we believe that the future is exciting at Vodafone and we believe that we in this room as telcos and using big data have the power to change the world and to do well by doing good thank you thank you very much louise put this very insightful presentation now I would like to invite miss Esperanza magpantai who is the senior statistician at ICT data and statistics division at ITU to deliver her presentation you have 10 minutes Esperanza thank you very much good afternoon everyone so the presentation that we just heard provided a very good example on the use of big data by the operators and I think this is not just the case in Vodafone but also in operators in our in your country or back home but for this session here what we what I would like to show you is a project that the ITU has recently launched it was launched last year and at this occasion at the same occasion of WTIS we mentioned this project as an introduction of what we are doing but today what we will show you is the outcome of this pilot project and we will hear experiences from individual countries who are with us today so the ITU as a UN specialized agency for telecommunication has a very unique position in terms of the use of big data because the telecom operators and a service providers are within our network so most of the work that we are doing in this area has been already discussed and presented in several occasions first of which was in the production of the measuring the information society report where we included a section or analysis on the use of big data and possible use of big data we had also included sessions on big data in the WTIS particularly the launch of this project in WTIS that was held last year in Botswana and we had included different discussions in our expert groups the ITU has been also a very active member of the UN global working group on big data and of particular interest to us is the use of big data that are coming from mobile operators and service providers so in June 2016 we launched this project which includes six countries each country representing the IT region so we have Colombia for the Americas Georgia from CIS Kenya for Africa Philippines from Asia and Pacific Sweden for Europe and United Arab Emirates for the Arab States so these countries have started the discussions on big data and particularly the pilot hopes to achieve a certain goal by the end of this project and particularly to come up with a methodology that can be used by countries in their data collection if they want to also look at using big data for enhancing information society measurements so for example in terms of the data that we collect from household statistics and those that are coming from surveys we see that a number of countries are still not producing indicators particularly for example on the number of the percentage of population using the internet so in this project we are hoping that we can have an indicator that can help us at least provide an answer into the the magnitude of this population especially for countries who are not able to produce data that are coming from household surveys knowing that household surveys can be costly it can also be time-consuming and it can be in some cases cannot be conducted regularly we also hope to see whether we can replace some of the administrative data that we collect using our traditional sources that are collected through the surveys that we send to the telecom regulators and ministries and finally what we want to to see is whether we can produce a new indicator that could be identified from this big data project so the different stakeholders in each of these six countries includes the telecom regulator the telecom ministry the national statistics office the telecom service provider so the operators and and the isps as well as the data protection agency so i'm not going to give a very detailed overview of the different roles because this will be emphasized by the different country experiences but this four key stakeholders play a big role in terms of the implementation of this project in each of these countries and from the side of the ITU we help we help guide these countries in terms of the implementation by providing assistance coming from our consultants working on us data scientists so we we help in terms of the processing of the data so the project was launched in 2016 and we are now at the final stage so this month we are having the final results coming from countries we have already have some of these results available but we are just at the process of editing so that it can be released to the public so at the end of the project that we are also hoping to have a final report where we will be able to compare the different experiences and results that are coming from these pilot countries so the big data indicators that are part of this project includes 16 or 15 key indicators which we hope that countries the pilot countries could produce and at the same time we provided them an opportunity to also calculate and derive the indicators that could be used for other purposes in their country so for example one of the the data demands that are coming from not just from the local but also from the international statistical data needs is this the SDG monitoring work so a number of indicators that are included in the SDG monitoring framework cannot be produced by just traditional indicators alone so we are hoping that this exercise can also be used by countries by national statistical systems so that they can learn on how to use big data for other purposes not just on ICT but also in other in other sectors so these indicators that I mentioned are included in a methodology document this is the main product that is coming from this pilot project and this methodology document includes description of the 15 indicators along with their processing methodology just in case countries wants to produce data coming from big data sources we also included some examples that can be used in terms of what are the expected results that could be calculated from big data sources and finally we also have some information on how data could be aggregated or disaggregated from the different sources this methodology document is was being tested in the six pilot countries and we were able to amend some of the initial proposals in terms of the the processing and the definitions and data sources that we outlined initially in the methodology document when we started the process so out of the six countries five countries use data that are coming from telecom service providers operators and ISPs so Sweden is the only country where we use a different data source and the result of the Sweden big data pilot will be presented in the EGT session which will be tomorrow so of the five countries Colombia managed to calculate around 14 indicators Georgia around 14 indicators Kenya nine Philippines nine in the beginning and I think they will report around eight indicators and UAE 11 indicators so there are limitations in terms of the the indicators that can be calculated from the different pilot countries and this will be elaborated to you later on why these are the indicators that they manage to calculate from their sources so one of the main results of this project is the the experience experiences learned and the issues that were faced by countries when they were implementing the project and access to data especially the administrative and legal procedures and documents remain to be the most important challenge faced by all the pilot countries so this is an area that should be addressed in in countries who are planning to embark on this new data new data source and it can be mitigated by different initiatives or different documents that each one of these pilot countries will also mention later on there's also some challenge in terms of the participation of the stakeholders so not all of the operators in the pilot countries managed to join the process and this is still related to the first challenge which is on the access to the data resources is a main issue partly because this is a new area of measurement big data skills are not necessarily available in countries in national statistical offices or in ministries and regulators so this is a challenge that needs to be addressed particularly to capacity building and resource sharing probably by countries who had already tested the big data work there's also some infrastructure issue in terms of how the data could be processed and what are the necessary infrastructures and application softwares that should be used to process the data in terms of the data processing model and this is related to the data access issue there are two basically two models that were proposed in this project the first model includes having the data being processed in the local vicinity of the telecom operators so if you have two or three operators they will be the ones each one of them will be processing the operators in their own premises and only then the aggregated data can be transferred to another agency in some cases it can be the regulator the ministry or the national statistical offices and this model was used in three of the pilot countries namely in Kenya, Philippines and the UAE the second processing option is that the telecom operators or service providers provide the raw data to the telecom regulator ministry or the national statistical offices so the data physically leaves their premises and it was sent to the agencies for them to process the data in their own vicinity and this was the case in Georgia in Colombia and also for the cases of Sweden so there are several deliverables out of this pilot project one of the the results that we are supposed to do is the presentation of the results of this pilot some of them we will hear today the second one is on the methodology document which we hope to make available after the editing work in December the country reports also will be available publicly online in December and the final report also in December because all of this are now with the editors and all of these output documents will be made available through a dedicated big data website that we have at the ITU I added there the link so you can check towards the this month that I mentioned so you can see the the documents that are expected from this pilot so if a country or countries wants to go and implement the big data pilot in their own country and we hope that this second phase we will have more countries joining the project we are advising them and based on the experiences that were gained from the six pilot that they prepare and solve administrative and legal documents that are necessary to access the data from their data providers so it is important that the close dialogue between the data providers and the national stakeholders meaning the ministry the regulator and the national statistical offices that they convene a meeting and talk on what data are needed and why it is important for them to share the data so like our previous presenter has mentioned big data for social good so part of this discussion among the different stakeholders is in terms of agreeing on the processing model for the data calculation on whether the data will leave the operator's premises or whether it will be processed within the premises of the service providers and then when in cases that the data will be analyzed in other agencies the mode of data transfer should be also discussed because this this require resources and infrastructure from both sides there's also a need of course to have a very standard and clear methodology and this is what we hope we have done through the methodology that will be an outcome of this project we hope that we achieve to have a very detailed description of the indicators so that when the indicators are processed by the operators themselves they will be able to follow it based on just the the different information that are contained in the methodology document we are also hoping to have some examples of algorithms that could be used by countries if they want to calculate it but the phone mentioned that they have this experience already that they have the codes that could be run in the countries that are in the locations that that want to use the processes that they've done and we hope also that this can be done in other countries that wants to to process the data and the most important challenge and we hope that this can also be solved and look at before our country or countries can embark on this project is the availability of infrastructure and human resources that will be used or are necessary when calculating or analyzing big data from different sources and of course last one is related again to the administrative and access to the data is the coordination and coordination and and and discussing with all the data data providers and all stakeholders so this is also to the providers for them to understand why it is important to release the data and to share the data and for the data users to be able to understand what are available and what could be used so that's all i'm ending my presentation with a with a picture of the methodology document and we hope that the next presentations will give us more insights in terms of the different experiences that were learned from this pilot project thank you very much thank you esperanza for giving us these comprehensive overview of this very important project as we know countries still face a major gaps in strategic production and i believe that this project is really very important and also for highlighting the challenges and the benefits that we can take from the using alternative big data sources now it's time to listen to three out of the six pilot countries how they experience this data production using big data and i would like to invite mr mohammed arlie who is the executive director of the national statistics and data sector federal competitiveness and statistics authority of the united arab emirates uh mohammed you have the floor and uh you have 10 minutes for your presentation thanks lately the international statistic community recognized the importance of looking at new data sources such as big data especially and the telecommunication industries because it produce a huge amount of data uh and informations and this can't be ignored by any statistical office especially at the era of digitalizations and as mentioned in the first preliminary sex sessions how we can deliver information or data first which is the real time data for policymaker and decision makers uh this pilot project by the itu honestly give us a good opportunities to look at the resources or the new data resources that we have in the countries and develop the methodology which has been presented or the document which has been presented to you just now uh UAE as everyone knows has participated one of the seven countries on this project and this project started with a collaboration between the regulatory authorities of the UAE and the itu with our telecommunication service providers and itu salat and do introducing the big data in ICT created and opportunities and this slide is being presented just now which give us an overview how to or or kind of a study how we could replace or complement the household survey questions and at the same times look at how we could replace some of the administrative record in our case also we were able to identify one new indicators we we've selected the option one where the telecommunications extracted the data in tier one and worked as an initial aggregated in tier two then it moved to the premises of the regulatory authorities where they work with the data scientists to extract the resulting indicators to summarize the UAE pilot there were five partners worked in this project and the collaboration was really important in this project with 44 trillions of event as initial raw data consolidated to one million data records with the data composed of the CDR and the IPDR with 100 coverage of the UAE subscriber and that was really a huge achievement for us 13 as mentioned by splancer 13 indicator being enhanced in our case and two indicators were not able to measure due to some of the missing information due at the time of the calculation and one in you indicators being introduced in our case using the big data analytics provided a holistic view of the infrastructure readiness of the country and the same times the people behavior in the countries and how they use technology in many others such information provided the governments potential indicators how the infrastructure is ready and the same time as the people people ready to adopt a new initiatives such as smart cities or internet of things which could be implemented in the countries the hundred percent subscriber coverage give us an opportunity to create a new indicators which is the origin destination matrix and this is this single data set provided a really comprehensive informations of the human mobility and the seasonality of their mobility also it covers the tourism movement who is using the roaming surfaces and imagine as a statistical office we always think also outside the area of ic tv so imagine if you could inject other dimensions such as or a human purchase behavior in the country so we know where actually people move versus where they actually spend the money and it could actually head different sectors like tourism economics and many others we did a mapping exercise to look at on the result to compare the outcomes and we've noticed that there is an enhancement and complement on the household survey data and also on a few of the administrative data and there is a potential replacement also from the household survey indicators in this case and in our case also it might require further analysis on the result because we still see some differences between the administrative data or the data that we have in the NSO versus the data that is being produced by the public sectors sorry by the big data there are a few challenges faced by during the pilot project which is mean which is mentioned which has been mentioned in the previous presentation but I have to mention here that maybe we did not have an issue with the resources our telecommunication which is Salatan do had really expertise in the big data and they had a group of data scientists similar to Vodafone and that helped us a lot to produce the data in much faster way as mentioned also early on in the presentation we had a quick we had a bit of issue in the administrative or the legality or the agreement to be signs to ensure the confidentiality of the information which has been extracted limitation or limitation and access to the data that was also one of the challenges unification of the data format especially between different service providers and developing the methodology and algorithms if we could if it could be provided earlier that will be really helpful and now with the methodology document I think it will be much easier for country to produce the data much faster one of the most important success factors in our case was the guardian of the regulatory authorities and the way that they manage the collaborations between all the stakeholders and the telecommunications to ensure and manage the confidentiality of the information that is being extracted and the result still would require further alteration to ensure that to ensure to to unleash the full potential of the big data because there is much more we could do with the data from the telecommunication industries as mentioned also earlier the developing the human resources very important in the area of data scientists or big data which will add a huge value to produce more analytics and finding an ICT industries developing an automation tools to pre-process the acquired data that will help a lot on the producing the analytical part of all the data received and also develop a new model which will protect the information and confidentiality is either by recording some of the confidential information of the of the of the users and the telecom or their or their customers that's and that's in general some of the lesson learns and this pilot to project with the ITU. Thank you very much. Thank you very much Mr. Mohamed Ahli for sharing this experience with us and especially on the lessons learned which is really very important in this context. Now I would like to give the floor to Ms. Lana Ramos who is the division chief of the program monitoring evaluation and statistics coordination at the department of information and communications technology at the Philippines. Lana you have 10 minutes you have the floor. So good afternoon everyone and on behalf of the department of information and communications technology of the Philippines I would like to share with you the experience of the Philippines as it as it participated in the big data pilot study of the ITU and before I start the presentation I'd like to say that I'd like to thank the ITU for inviting us to participate it was indeed a very it is a very educational experience and I would like to highlight some of the important issues and challenges that we encountered during the project. Well first of all as it was already explained earlier we were invited to participate in the big data big data sorry pilot study during the ITU conference of May 2016 and as explained earlier this is a collaborative study of six countries and the department of ICT was identified as the agency focal for the project. It is ongoing as we await further results data results from our providers. Jumping to our project resources some of our data providers found it essential to supplement their current teams for the project an average of four to five members for each team was required a dedicated team actually had to be formed for the project as it entailed additional work and resources for the project. For the ICT as it's designated focal we also designated a specific team to oversee the project and the support of the legal division was very integral as I will explain early in later slides as legal aspects were very important. The National Privacy Commission which is an attached agency of the DICT is also a very important partner in this project and it oversaw privacy concerns that arose. The advantage for the Philippines well the ICT sector is one of the sectors that is enjoying a very positive trend and we see big data as contributing to further growth of the sector. It also has an active mobile phone market and right now we count more than 120 million mobile cellular subscribers as of December 2016 so the use of big data is really going to be a potential contributor for development and policy making. The DICT also sees the project as important in paving the way for how we will work with private telecommunications companies. So the pilot project essentially laid the foundation perhaps for future public-private partnerships in terms of data sharing and maybe engaging the private sector in more active public sector activities especially as we move towards big data planning and policy making. It was also an opportunity to test the new legal environment for data protection. The project happened to come at the height of the environment in which we were trying to test the Data Privacy Act and its implementing rules and regulations so it was a good opportunity to see how these regulations would impact on data privacy. Our stakeholders were two very big telecommunications companies which account for about 90 to 95 percent of the market so they cover a majority of the telecommunications market in the Philippines. They are our two very important private partners for this project and as I mentioned earlier the National Privacy Commission is an important partner as well and the National Telecommunications Commission. Keep milestones. These are important activities that mark the contact of the project and the project kickoff was held in June of 2016. February we held a threshold analysis briefing with regard to the contact of a privacy impact assessment study to look at privacy issues that the project raised. In March we had several technical meetings with ITU. In April we hosted the visit of the ITU consultant who came to oversee the data processing and in July we welcomed the first submission of the data from our providers. Concerns and actions taken. These I think are very important issues that should be discussed as we move forward perhaps in a phase two of the project and as the DICT itself moves forward with its own projects on big data. So privacy concerns and access to the data. Primarily this was a major concern raised by private sector companies and we addressed them through developing legal instruments such as a memorandum of agreement that covered roles and responsibilities of data providers and the DICT. We conducted the privacy impact assessment. This was a very important step as stipulated by the National Privacy Commission. This was in fact non-negotiable before we could move forward with the project. And the DICT achieved most of the risks in the absence of NDAs between data providers and consultants. So the key document here as far as the Philippines was concerned was having a very, very comprehensive memorandum of agreement. With regard to the indicators as mentioned earlier we could only produce eight of the 15 required indicators and this was because of the risk of exposing sensitive data as raised by our providers such as antenna locations and this impacted on two indicators. Unavailable data and time constraints to process and this impacted on one, two, three, four, five, five indicators. And so we ended up with the final eight indicators BDO3, BDO4, 5, 6, 7, 8, 9 and 11. Data sharing and transfer again caused some issues. The issues were mainly which way we could send the data that could ensure that this could be as private as possible, as secure as possible. And we ended up with deciding on using one of the DICT's file transfer protocols and this is called paquete.gov.ph and this was used by one of the providers in transferring the data. This remains to be some of the issues moving forward, how to secure data transfers. The memorandum of agreement, if I may highlight, had relevant portions which many of you may want to consider. The appointment of a data protection officer was stipulated in our memorandum of agreement and this is on the part of both the private sector and the public sector which was the ICT. The contact of a full privacy impact assessment, again this is part of the implementing rules and regulations now of the recently passed Data Privacy Act and the data to be provided as anonymized so no trace of personally identifiable information should be submitted for the project. Another risk mitigating measure that we adopted was the process data had been processed within the premises of the data provider for the project. Finally, some recommendations moving forward in the case of phase two perhaps. A risk analysis for each country to include preliminary study of privacy and data protection concerns. This is very important as this caused some delay in the case of the Philippines, several months in fact just to ensure that all privacy concerns were addressed. Pre-project consultations with target stakeholders on the commitment and data they are willing to share. This is again important because you would like to get the full commitment of your data providers as well as the data and level of data that they are willing to share. Development of legal instruments to define roles and responsibilities and data management cycles. Again these were the key documents that provided the foundation for the actual implementation of the project so this is important from the very beginning. This should be ensured and securing data sharing processes and platforms. Again from the very start we would recommend that data sharing and platforms to be used should already be discussed as early as possible and that would conclude my presentation. Thank you very much. Thank you very much Ms. Ramos for sharing the Philippines experience with this pilot and mainly for highlighting the importance of the memorandum of understanding and also all the privacy issues around this data sharing. Well now we will move for our last presenter Mr. Juan Davi Orlate Torres for presenting the Colombian experience. Mr. Torres is the head of the planning sectoral studies from the Ministry of Information and Communications Technology in Colombia. Mr. Juan you have ten minutes on the floor. La oportunidad de generar este proyecto piloto en Colombia de aplicar el modelo Big Data para la generación de estadísticas. En Colombia realmente le estamos apostando a la generación de estadísticas a través de nuevas fuentes de información y el aceptar digamos este reto hace que las condiciones o permite que las condiciones en Colombia hayan tenido las dimensiones necesarias para poder desarrollar este ejercicio de la mejor manera hemos venido trabajando fuertemente en la generación de estadísticas primeramente via registros administrativos eso ha permitido que se genere como una relación de confianza con los operadores este ejercicio lleva más o menos ya cerca de ocho nueve años y aparte de eso se viene complementando también con la generación de información a través de fuentes primarias a través de grandes encuestas en materia tick en este caso en particular este sin duda era el nuevo paso nuevo reto Colombia efectivamente le está apostando a la generación de estadísticas a través de una nueva fuente de información y por eso nos cayó como se dice en Colombia comanillo al dedo este reto en Colombia trabajamos a este proyecto con tres entidades básicamente a través del ministerio de las tecnologías de la información y las comunicaciones quien yo represento que es la que se encarga la generación de la política pública en materia de tick está nuestro departamento nacional de estadística que es la cabeza rectora o el interrector en materia de estadísticas en Colombia y el operador claro que es uno de los tres operadores de telefonía móvil celular en Colombia pues nosotros fuimos el punto de contacto me refiero al ministerio tick punto de contacto con la itu y nuestro departamento estadístico fue el validador y fue el que nos ayudó a articular al interior del país la generación de este proyecto de vitata cuáles fueron los principales acciones o la descripción de lo que hicimos inicialmente mandamos o enviamos tres comunicaciones a los tres operadores en Colombia que operan en Colombia como tigo movistar y claro claro fue la entidad o fue el operador que nos respondió afirmativamente como también lo hizo en su momento movistar pero más adelante les comento movistar se inclinó por otro modelo y con otras condiciones que nos impidió en ese momento digamos continuar con con movistar y por eso nos fuimos con el operador claro e additionalmente a ello tuvimos que mirar la información que por ahí se iba a generar como es información bastante sensible para los operadores nos tocó así al igual que nuestros países que también desarrollan este proyecto un acuerdo de confidencialidad porque sin duda los operadores son temerosos de que eso llegue a afectar su negocio como les decía entonces el operador claro fue el que nos propuso desarrollar este ejercicio conjuntamente con la metología dos información que significa la metología dos o las el segundo segundo modelo y es que ellos compartían la información y nosotros internamente la procesábamos y realmente claro en colombia tiene una porción del mercado del 50 por ciento y eso pues nos daba digamos una información mucho más amplia que pues nos daba entonces más información y mucho más precisa de lo que nosotros estábamos buscando y después de ciertos retrasos algunas inquietudes que resolver que más adelante también las se les voy a presentar en los temas de desafíos iniciamos entonces este reto en particular de conseguir la información cuáles fueron nuestros recursos básicamente fueron 10 personas que estuvieron envueltas o que estuvieron trabajando en este ejercicio gente de ministerio tick gente del departamento estadístico nacional y gente obviamente el operador claro con todo el acompañamiento de los de la itu nosotros nos tratamos cerca de un año y dos meses en desarrollar en procesar toda esta información de este año y dos meses mucho más que la mitad de este tiempo se tardó o se explica por la dificultad de firmar esos acuerdos de confidencialidad la infraestructura pues contamos con cuatro nuevos virtuales de los cuales hubo cinco terabytes para funcionamiento 24 gigas de ram y un 8 cores digamos de procesamiento trabajamos a través del sistema operativo oracle linux bueno igualmente estos datos estos indicadores se calcularon con base en las herramientas map reduce y spark y fueron corridos en la plataforma hadu cuáles fueron los principales desafíos que nosotros encontramos cómo convencer al operador para que nos pudiera suministrar la información y pues siempre tuvimos como como premisa tener un gana gana decir nosotros ganamos teniendo la información de los operadores para poderlo obviamente estudiar y mejorar nuestra política pública y obviamente el operador tenía también que ganar algo y lo que encontramos en el ejercicio de este gana gana es que el operador nos mostró su interés por dos razones fundamentales la primera es que le generaba ellos un reto técnico o tecnológico el poder procesar toda esta información y la segunda que tal vez es la más llamativa es que ellos solamente hasta desarrollar este proyecto se dieron cuenta la cantidad de información que ellos podían generar y que no las estaban utilizando esto sin duda generó digamos un interés por parte del operador para proveer la información igualmente desafío relacionado con riesgos no previstos por ejemplo todo este debate de la anonimización de la información privada eso fue lo que generó mucho más debate y retrasos en el ejercicio de nuestro proyecto los recursos tecnológicos en el ministerio no teníamos la infraestructura tecnológica para poder procesar la información pero si contamos aquí con un aliado especial que fue nuestro departamento estadístico que sí contaba con una infraestructura robusta que nos permitía desarrollar este ejercicio sin duda ese es un tema que hay que tener muy en cuenta y también el tema de reprocesamiento permanentemente nos tocaba validar con el operador porque la información que salía inicialmente pues no era consistente con los datos administrativos que tenía la misma operador ministerio unos poquitos indicadores o resultados nosotros efectivamente trabajamos los 14 indicadores y vamos a trajimos para conocimiento de ustedes cerca de cuatro indicadores da cuenta de el uso de la tecnología o el uso más bien de las actividades más sociadas con con temas de internet o servicios de internet por ejemplo temas de mensaje y de texto y de voz lo que podemos encontrar aquí en el mapa de colombia es que efectivamente la tecnología más usada la 3g y en los lugares donde están digamos más marginados digamos de la periferia que están en la periferia pues la tecnología que más se usa las dos g y obviamente donde está concentrado la grandes poblaciones es donde se empieza a ver un desarrollo más amplio de cuatro g cuando ya vemos en el asunto en temas de tecnologías para uso de internet pues la participación sigue siendo alta la doge pero tres g digamos sin duda la que mayor participación tiene en el mercado pero empieza digamos a irrumpir la tecnología cuatro g con cuatro millones y medio digamos de conexiones en colombia la tecnología cuatro g es la que más crece y eso digamos se va registrado en este cuadro venimos creciendo más o menos trimestre cerca del 15 por ciento crecemos en ese uso de tecnología y además se va a haber complementado cuando estemos subastando la banda de 700 megahertz también tenemos otro indicador relacionado con el rooming cuando lo activan los extranjeros cuando llegan a nuestro país recuerden que este solamente son los resultados del operador claro que más o menos tiene la 50 por ciento de la participación del mercado y aquí podemos registrar que sólo con el operador claro cerca del 22 25 por ciento de los extranjeros que llegan a colombia activan the rooming si nosotros adicionamos los otros dos operadores pues muy seguramente tendremos el 50 por ciento es decir la mitad de los extranjeros que llegan a colombia activan el rooming finalmente otro indicador que quisimos mirar es el desarrollo de la infraestructura móvil con el asociado con la población y lo que ustedes pueden observar es que efectivamente donde está concentrada la población es donde está desarrollada mala infraestructura todos estos resultados sin duda lo que nos va a permitir a nosotros entre otras es mirar cómo podemos atacar o reducir las brechas digitales colombia es muy marcado a veces esas brechas y sin duda este ejercicio de generación de estadísticas a través de big data pues nos va a permitir sin sin duda afinar nuestra política pública en este particular las recomendaciones finales el tema jurídico es lo que más toma tiempo en nuestro caso como les decía seis siete casi ocho meses duramos en el debate jurídico para afirmar ese acuerdo de confidencialidad fue bastante desgastante pero finalmente pues se logró definir la hoja de ruta una hoja de ruta clara donde no solamente a borde los temas de la infraestructura tecnológica sino también temas de procesamiento donde está incluido obviamente los científicos de datos que en estos casos son tan importantes y que en el caso de colombia pues fue proveído por no solamente por el acompañamiento de la vtc no también digamos con por a través de el nuestro departamento nacional de estadística permanentemente hay que estar revisando los datos que se procesan para no esperar hasta el final y volver a tener que reprocesarlos es bueno irlo permanentemente trabajando en la media que va saliendo los resultados igual de igual manera poder trabajar los resultados a nivel micro para después poderlos expandir a nivel macro y finalmente estar validando la información que se que se viene procesando permanentemente esto es básicamente los digamos los resultados del proyecto de big data en colombia y pues con esto cierro la presentación muchas gracias muchas gracias one for sharing your country experience in this pilot has been part of the statistical community i think that i can say that we acknowledge the benefits of big data sources for the production of official statistics statistics and i think that what i saw from your experience is that besides the technical issues that must be addressed and also methodological issues we have to look very careful carefully how to engage stakeholders how to design proper agreements that treats privacy confidentiality and so data sharing is really key in this process and i think that from this i2 pilot experience we can learn a lot and how to scale up to other indicators we still have about 10 minutes and i would like to open up the floor for questions and comments the floor is open yeah i guess so no questions yes i cannot read your please can you identify yourself please and then tunisia yeah tunisia you have the floor merci monsieur le moderateur mesdames monsieur les panelistes je vous remercie pour la qualité et la clarté de vos interventions il est évident que les pays émergents devraient être impliqués dans l'utilisation des big data en tant que vecteur de transformation dans la société de l'information à ce titre nous proposons de considérer l'évaluation de la fiabilité des résultats d'une étude statistique comme un nouvel indicateur nous considérons que les big data contribuent à l'amélioration de cette fiabilité à travers la fourniture de données opportunes plus fréquentes et plus granulaires une étude statistique fiable impacte sans doute le processus de mesure et d'évaluation de la société de l'information d'où l'importance de ce nouvel indicateur la tunisie souhait aussi considérer la croissance résultante de l'adoption du big data en entreprise il s'agit de mesurer le chiffre d'affaires que devrait générer le marché de big data cette mesure peut considérer principalement un les nouveaux canaux de marketing de les nouveaux flux de revenu généré par la vente de matériel logiciel et produit complémentaire et enfin le nombre d'emplois généré sur le marché et merci thank you tunisie we have another one yes please you have the floor thank you mr. moderator and mr. chair may i provide an additional input on the presentation of mr. Alana Ramos on behalf of the department of ICT of the Philippines we shall continue with data in a big data initiatives even after the completion of the big data pilot study no it is because we see big data as important in policy formulation and in implementation of the free internet a wi-fi across the country now being in a deregulated telecommunications industry and based on what we have experienced in the conduct of the pilot study we aim to strengthen our partnerships with the telecommunications operators and ISPs in identifying ICT related big data now we also aim to partner with other private big data providers so that we are able to capture relevant big data for us to be able to develop grounded and evidence-based policies now this data providers could be from the private entities or academic organizations which provide open source applications now this is like what the bandwidth signal and statistics are the best project does now I understand this will be the showcase in a discussion this afternoon as well now once more in behalf of the department of ICT of the government of the Philippines we take this opportunity to extend the gratitude for including us in the ITU pilot study on big data now we hope that the Philippines again be part of future ITU initiatives thank you mr. chair mr. moderator thank you very much the Philippines Bangladesh please no yes please you have the floor I'm sure I'll be able to talk about the other thing that I want to say is that there is no one international presence until now the people working on it like this the third thing is that it has not even now the system and solutions and the programs for all the people with each other I remember that the three challenges that you are facing are facing thank you Mr. Dome the last question from Brazil and then I will give one minute to each speaker Brazil we have the floor hi thank you mr. chairman I'm Daniel from the regulator of Brazil we had a similar study in Brazil we tried to conduct a similar study in Brazil during the World Cup and the Olympics we worked together with our NCO to collect data on roaming during the event but we faced issues like we as regulators we didn't have the processing power to process the information or the storage capacity and the operators they didn't comply with the restrictions with the amount of the investment they needed to to process the data themselves and to storage the data so we had to drop the whole project all together because of these these issues that other currencies are also facing thank you thank you Brazil the very last intervention please good afternoon and thank you so much for the presentations was very illuminating I noticed that in a lot of the countries that you had to agree some mo use with the operators I wanted I wanted to know the you know the major content of those mo use well what were the issues that operators were concerned with secondly I also wanted to know the cost of the project who bought the cost was it the operators I bought the cost or it was a combination of both operator and government thank you thank you very much now I would like to give one minute to each speaker to make your final remarks and maybe address in some of the issues in particular the last one related to the memorandum of understanding the agreements and the costs and all the technical issues so I'll give one minute for each and I will start with Louis thank you it was very interesting to hear the challenges that you're facing I think the closer you can keep your privacy officers the better and I think all of us have worked to do in making big data seem less scary I actually personally hate the term big data I'm sure it was great when it was coined by whatever consultant to helped all of us in the room get salary increases by calling ourselves big data analysts but the the public perception now of big data makes it seem very scary intrusive internally we've pivoted more to talking about data analytics because what we're doing is trying to improve the customer experience first and foremost the more we can talk about big data in simple language the more we can be transparent the more we can show very directly the benefits to consumers the better this will be for all of us so make friends with your privacy officers have them sit with you as they do for me and that's my best practical suggestion on how to get started in in making this an easier process for everyone thank you Louise Esperanza you have one minute thank you very much I would like to just probably address the issue of mo use and mda so what we are also hoping to do in addition to the indicators definitions and methodologies so the document that we will produce as part of this project will also include examples of mo use and mda's that came out from this pilot so that those that wants to use them in the national context would be able to adapt them according to the needs of their stakeholders so that's one thing and at the same time on the applications and and softwares that are what should be used in countries I think out of this pilot there are not necessarily a standard application that could be used but I invite you to also talk to this pilot countries and see what would be best fitting to your needs and at the same time what are available in in your operators already because in some cases you may have already those applications that exist in operators that you don't necessarily need to to reinvent the will thank you thank you Esperanza Mohammed you have the floor maybe I have two comments here number one is that the data the telecommunication it's could be used by the service provider to commercialize it and marketing and marketing so you have to have to include that in your memorandum of understanding not to not to break their maybe one channel of business the other point is the methodology for big data and statistics how to calculate the indicators is very clear and it has an international agreement between all the nsos how to produce the indicator related to statistic so the same things I think this exercise with the icts if the methodology is being agreed internationally discussed with the statistical offices and do a comparison of the result where at least both result of both results statistically and from a big data kind of match then I think we we would be able to sort out or to come up with agreement on the methodology to be used for big data indicators thank you Mohammed please you have the floor yes thank you with regard to the memorandum of agreements as I mentioned this is very central this is very central to the project and maybe perhaps as lessons learned you could look at not just bilateral bilateral agreements but tripartite perhaps because it will minimize so many agreements so you can probably include everyone in one in one memorandum of agreement in in our case it's very specific to the Philippines so we were compliant with the data privacy act as well as implementing rules and regulations so a very key key component of the the MOA was having stipulating the the inclusion of a privacy impact assessment study as well as other key key provisions but that's what I can I can advise thank you thank you Alona this is indeed the multistakeholder effort it's not only bilateral agreements one you have one minute yes not only with that political will but also to work in various aspects that strengthen the information system of the country a robust statistical system with a very robust infrastructure that has credibility and that accompanies all these statistics generation exercises the generation of confidence with the operators all let's say on the table donde se puedan ver realmente cuáles son los intereses de cada parte y donde se les pueda evidenciar lo que sobre todo para para el operador lo que ellos pueden ganar con la generación de estas estadísticas y sin duda por detrás todo un equipo técnico riguroso de las partes involucradas que permitan tener pues la certeza y la confiabilidad en la información que se que se procesa básicamente serían las las recomendaciones finales thank you one thank you to all panelists I think that just a final remarks his parents I have mentioned the UN global working group on big data for official statistics and I just would like to recall that this ITU initiative is very much aligned with the initiatives that is stated in the Cape Town the UN Cape Town global action plan for sustainable development data that puts a lot of emphasis on revising the fundamental principles of official statistics and on the need to embrace open data initiatives as well as to remove barriers to the use of new data sources such as big data so I think that it is our responsibility to foster the debate in our countries and discuss how to engage stakeholders government and so private sector around this issue on use and sharing big data source for official statistics and with that I think that our member states are really grateful for the ITU initiative in promoting this pilot project especially I would like to thank the ICT data and statistics division that has put a lot of efforts in helping and assisting countries the six countries in trying to take advantage of these big data sources and try to develop new methodologies and I think that as I said as part of the statistics community we have to embrace this open data initiative and innovation and modernization of national statistical systems depends highly depends on all these new partnerships this mouth stakeholder approach and stakeholder engagement we still have a lot of the technical issues methodologically to be addressed but we have to foster this debate in our countries with that I would like to thank all the speakers for the very very insightful presentations and I hand over the floor to Mr. Chair thank you very much thank you very much for all panelists and it was very useful and deep and I think we learn a lot from this experience and trials as I said I think one year ago I said in this place that our petrol is the data and as we learn together to develop the petrol industry and to get diesel to get plastic to get gas and to get different sector of petroleum industry we have to build up together the data industry and this is a long process that we are developing together and I believe the experience of ITU is very useful in this state for this session thank you very much and see you tomorrow thank you