enrichment for the data and so the data from data sf, comes with latin long, associated with it and latitude and longitude coordinate and so for a contributor, you can see where the contribution was made and that is useful, but we might want to know from a lat long, what state was that lat long from? what sf district was it from. what california county did the money come from? and you use that, you know, enrichment, to show further correlations between things, do we find that certain kinds of activities are associated with certain geographic regions. and you know, to answer that kind of question, requires, some sort of processing. and processing a data set for that geo enrichment can take, hours to days to find the positions within these geographic regions. and so, there is a lot of processing that needs to be done, and there is a lot of sort of intermediate data that needs to be stored, and going forward we are going to be looking at more of that, and seeing if we can make more out of sort of this foundation, that has been provided to us. and so, yeah, we are very excited and one o

wide data standard.fully and effective data act will address that issue. understand goa is working with inspectors general to develop some of the standards so that there is auto oversight protocols to monitor compliance with the data acted. that might be something that comes up today as well. finally i would just want it say that this is a great example of what we can do when we put our minds to something that helps our tax payers who we represent. it was both branches so bicamral. here on the hill and with the executive branch. bipartisan. i would say even nop partisan. and a very difficult within let's face it. it was decided, here so problem, let's attack it p.. and there is murky information. members of congress from both parties engage each other. engage the executive branch. we hammered out the details with a solution. by the way, as objections emerged, we work together to try to address them without waivering in addressing the problem. this may not apply to every issue but there is certainly a lot

with that april data released. we will not dwell what was on the april release but what could be in future vises but they can talk about what they would like juicy to make them more useful. but it is not the only game in town. and he can go through the process to save time through the data use agreements to get the data that is behind the release. to end either can work with the entity and also has another channel. >> but the first question with the charge of pavement is the big deal from the perspective of the researcher with had a couple of conversations and then salivating over the new data source. and gradually there is a much we could do with the data the research questions we would want more detailed micro levels then redo all the time. sodas that the data releases not the big deal? not all. is the step forward. in number one of the release of the data that cms says the physician's right to privacy does not outweigh our right as a public to know what medicare is spending or buying. and then in place and to be

data. if i was a board member and you and i were back in our previous lives, we would say, i would say how can i say okay to more resources, and more effort, more taxpayer involvement, when i don't know the return on investment for the cap-ex, and i don't know what the operational effectiveness is other than really macro data. can you shed any light on that? am i -- and am i missing the boat here? >> i'll shed light on my commitment to more transparency. i think part of the problem we have is lack of coherent data. and so one of the things i directed in the executive actions that we issued week before last is i directed the office of immigration statistics to create the capacity to collect, maintain and report to the secretary data reflecting the numbers of those apprehended, removed, returned or otherwise rerate rated by any component of dhs. i also intend that this data be part of a package of data released by dhs to the public annually. so, i'm sympathetic to what you're saying. and i wo

the ability to combine data and data sources and data sets to get an important view

data, that is information about the data as itself, canhe data be misused.he industry oppose the u.s. 215 stuff because although the government a misuse ofas not the data, our argument was that itself could be misused. i think these are learnings that nowcountry has to now forever. snowden, hero orville land, year and a half look? how does it >> it's interesting, depending this america you get a different answer. juliane same with assang. the eastlk to coasters, they tend to take negativengs in a very context. west coasters, again these are iteralizations, tend to view in a more civil libertarian view. and i'll let the court decide on questions. we clearly do not want to encourage bulk data leaking. imagine again, think of all the data bases that exist about us, taxof you, health record, record, where you are, what you're doing, it's not good for people tone courage do that. on the other hand, his helpful in were shining the light on these practices that people like myself and maybe you didn't know existed. possible.ey were but when it starts monitoring traf

you know, payment data was encrypted but in transit data wasn't. it's encrypted now.what do we have? we have ram malware. even after target, home depot knew about this. it was in the papers. they got hacked and didn't know about it from april to august. so, you know, andrew is now with the wireless company. you don't need to factor two awe they wancation -- authentication. in the '90s they wanted us -- wanted the government to prosecute the access devices. they got the legislation, that didn't do that, so what did the industry do? they enscripted the identifier, built a system and why if my phone is uniquely protected, better than me, why can't industry protect me and my identity? >> so how i'd like the answer that question is you'll start to see it in -- first in small pieces. and then a large -- a larger movement. so what to look for are technology companies who are opting not to collect data. and maybe it's the perception of extra work, there's no way to monetize it. the example that comes to my mind first is apple pay. one of the unique things about apple pay comp

benchmarks we'll be monitoring every quarter as best with the data and as we get the data coming to us frequently to keep track of the fatality injuries by neighborhood and age and make sure that is effective citywide and we've got a strong commitment that the public health department to help us with the sf medical with the hospital so see a trend whether up and down and an interim metrics looking at the speeds on the streets we've made huge progress in reducing speeds but we need to pay attention to the aerials and we want to see about the citation issues by violation types in the districts we'll see nor information it's been a challenge i'm not going to print a rosy picture but something to address we are tracking those and looking at from the da's office of investigation and prosecution of manslaughter we need to track that i don't know the outcomes but we're going to be focusing on that. >> can i ask point out on the per actually of speeds we're going to have a budget analyst report the limit reductions and efforts to reduce speed limits i know that is coming up on the end of janua

google has a very large private data network that moves the data around and complicated, powerful ways. that is why google works so well. it is a massive technical achievement. i read it in "the washington post." i was shocked because jerry: i had written a book that talks about this being possible and people have hinted that it was possible by monitoring light fiber, although i think the mechanisms did not apply. we did not know. the fact it had been so directly and documented in the documents that was leaked was really a shock to the company. here we are. we are more than a year later from that day. the morgan stanley high-tech 35 index, about 25 percent. google remains one of the most massively profitable companies in the world. >> has there been real damage to you or the industry or to us more broadly? >> there is damage at many levels. let's start by saying that if you are a european right now you are less likely to trust an american firm to take your data. maybe you should have all been -- always live in -- always been less trusting. i google, we massively encrypted are google se

that data very quickly.e see that when the industry was being attacked, we saw fishing and spaming increase immediately. so it's that having the infrastructure to permeate the crimes that you want to commit and know you're going to be able to monetize quickly. i don't think other groups are as organized as the russian-speaking bad guys that we deal with. >> my answer would be it depends on global economic shifts. why i think eastern europe is the center of cyber crime is partially network effect and the internet makes everyone a target globally so you can target someone across the world just as easy as you can target nearby. and also they don't have a silicon valley. so the reason silicon valley works is companies and talent. and so they have a similar business network there. so that is a draw although criminal and if there's a rich technology industry nearby for people with those interests that is even at lower pay will be a stronger draw than the criminal under world. >> a small story here. . six months ag

so it's about having people use the data and not just having the data available. by using a we are able to identify strengths and weaknesses of that data. so where we work where we have a group of scientists and doctors and consumers and writers who take data from various sources and translate this and we look at the areas of preventative services like cancer screenings and hospital quality, position qualities and also a health plan quality and so we have been doing this for about 10 years and we've reached about 20 million consumers per month. even more through our partners which are not just online, but often those that can reach other audiences. so you're probably more familiar with cars and electronics and we often compete with those areas and we make fun of each other. and so i'd be rather and hospitals and all that. [laughter] and so it's a healthy rivalry. but i was looking at data the other day and i was really excited to see that in our own measurement of how our readers use data and how we understand the stories that are stories on health care cost and q

data? for us we were grappling with as an auto industry we focused on data that is identifiable. if it's agzpwated or deidentified it was not part of the calculus. and then there are pieces of vehicle data collection that are going to be optional to you like if you want to be a probe on the wave system you can choose to be or if i want to wear a fit bit i can choose versus some systems where there's not a choice. i don't know if it's a useful distinction but it's one i've een focusing on. >> i think piggy backing that a little bit. a lot of it starts with the education that we talked a bit earlier that is in the report. people need to start understanding the amount of data they're generating. when i bought a fit bit i didn't think about all the different places that data would be. i don't really care who has the data on how many steps i take. but we did a report last year that looked at all of the data and all the different places and the different vulnerabilities that it goes through. agai

data. just in case any of you would rather use different media. online,re viewing us you can immediately download a couple of resources that i wanted to point you to. once i finish this housekeeping we will get to why we are all here today. the two items that we have that are available in the room, if you are here in person, are one, a first issue brief on the commonwealth website as well as ours regarding cms medicare physician payment data and the topic that we are discussing in part today. and then also something that i caitlin brandt, who helped with this effort, for, once we spoke to so many of you we learned that there were so many tools out there around transparency that even to us to talk about this on a daily basis it became overwhelming, so we decided to for the sake of today's conversation put together our own online nonexhaustive resource list. we would love feedback on this and we would encourage you to send us the act to the website or through the twitter feed so that we can g

physicians are very data hungry, interested in data. if they view it as credible, but they want the ability to do what they call drill down. to drill into the data in a way that makes sense to them. all of this is what i would call a service bureau function and the things united health care has done, as in our premium designation program, we've done since 2005, we have a physician portal that has hetus data on it open to all physicians in our network. this service bureau function is an important suggestion how to enhance things going forward. lastly, let me make a few suggestions for where i think this could go from here. i think i've said some of these already. first is to just clean up the data and offer some simple descriptive statistics for all stakeholders. i think that's a reasonable thing for the sponsor to do, even as others can do other kinds of analysis. second, i agree with your point, the ability to combine data, data sources to get an important view, in this particular case, it's really quite a limitation that you can't get

i think to some extent, it's a big data solution that helps solve the big data problem.the more data that's being generated to do -- that could be stolen, can also be leveraged to do, you know, analytics to protect us. so i think it's important to -- it could be for preventing fraud, it could be for authentication purposes. we make it so hard to do these kinds of things. how many people are using face pook? no one's going to admit. good. how many of you turned to facebook? it's hard. you don't even understand it. and i tried to explain that to my mother-in-law, she doesn't everyone want to talk about that stuff. we have to rely on the people that make the software to make it easier. >> we could use things like machine learning to identify your log-in pat eterns to say h, you usually don't log in from south korea at 3:00 in the morning. that seems pretty weird. why don't we ask you for a two-factor n/m"[ when any other time we don't. because you're logging in from home from your computer at normal times. >> first of all, i think judge posner would present economic analysi

they called it data stash and will let customers utilize unused data from prior months up to a year.t's part of the un-carrier strategy that includes phone financing, the removal of contract, and free music streaming. i spoke with john ledger and started by asking why he is making these changes. >> we are solving customer pain points. it they are not marketing programs. they are fundamental industry changes that are permanent. this is one of the most infuriating gigantic issues that customers have pointed out to me. 40,000 social media messages. at what it's about, the paid customers that exist, 110 million of them have teacher phones or unlimited which leaves 120 million people that have to purchase data. 90% of them have no idea or how much they need. what happens is the penalty of using too much. $1.5 billion in overage penalties. we do none of that. and the rest of these people, they purchase an average between four gigabytes and five gigabytes of data and at the end of the month, three are unused. that's about $12 a gigabyte. $15 billion a year. -- $50 billion a year. it's like

the data that's been released. is it primarily a question of consolidation? so you download less data to answer the same questions? is it primarily hipaa, is it some combination? >> yeah. if you go through the process of entering into a data use agreement with cms, you -- and you guarantee that you're going to satisfy all of the data security safeguards that they require which are serious, they will send you claims data that has a beneficiary id, that has all of the -- [inaudible] zip code of residence of the beneficiary that has the microlevel service of the cms physician public release only reports combinations of physicians and wpt codes -- cpt codes with 11 or more services provided. obviously, if you buy the micro level data, you can see all the services. but it's a major hurdle, to set up data security systems that are impenetrable and are hipaa compliant. >> and also provides the opportunity to pull together various services from different providers that a beneficiary has had. >> and you have to promise not

as the last official data. and i'm wondering if you have had had a chance to look at anything that might be more recent. and if so, could comment briefly on what it shows. >> right, 23009 was the -- data from 2009 is the last data that has been cited in official publications by the national highway traffic safety administration. however, those data are -- those statistics are based on data that are collected annually and presently are available through 2012 and the most recent data, which i have looked at from 2012, show that 1.6% of drivers involved in fatal crashes in 2012 were believed to be drowsy or fatigued at the time of the crash. >> what would you say is the general trend for the drowsy driving data? is there a general trend and if so what direction is it heading? >> well, in the official statistics, there doesn't seem to be much of a general trend. if anything, the proportion in the official statistics may have decreased slightly in recent years. however, as i talked about in my presentation, i believ

it's tight to where the instruments collect data. >> for scientist, every bit of data counts. >> youave to map all the area, even the areas that we think are not changing. getting a map of the ice sheet is important, and we almost do it one glacier at a time, that's the benefit of mission. >> i can't get over how beautiful that place was. glaziers? >> it was so unbelievable to fly at the low altitude. the aircraft is designed to fly close to the ocean. it was built to survey submarines , so it was designed for the job. but the views from the cockpit were stunning. i was pinching myself the whole time. we were looking outside the window at polar bear tracks. it was amazing. >> i understand the pilots can see a difference outside. what about the science behind that. there's a lot of equipment on the aircraft. they measure the bedrock under the ice sheet. they found something that rivals the grand canyon. we learn about that in part two. >> i magg managed to get into the cockpit which, for a mechanical engineer, is a dream come true glaziers. twitter. [ ♪ music >>> >>> hey, guys, welcom

data. as was mentioned, once they got together in pairs started putting the data and they said, we could actually use this for transparency purposes, start to look at quality data. as some of you may know, that is an effort that has been announced and will be deployed. to sum up to five and the other thing, a medicare qualified entity. that's a good example. so to so to sum up, i think, a lot of opportunity. this is the beginning of a knew era's. transparency is the coin of the realm. the current release has been useful in some respects and limited in others, but others, but i think it we will evolve and strengthen overtime. >> just one point you made that got me thinking. thinking. you were talking about one of the steps forward is to go from raw unanalyzed data to various levels of analyzed data. data. at the same time this is evolving we are on track to have physician value modifiers which we will be presumably fairly analyze data. i don't know what the transparency plans are, but any th

so we decided to add up violent crime data and property crime data to figure out which city really contains the most crime overall in the state. and it turns out emeryville is above and beyond the most crime ridden city in the state. its overall crime rate is twice that of the second and third most active cities eureka and oakland respectively. while oakland does grab the headline emeryville actually sees the most violent activity. for a look at your crime rate in your city head over to nbcbayarea.com slash reality check where we have compiled a more detailed list. now to say that oakland doesn't have a crime problem would be far from accurate. it came in third out of more than 350 cities in california. but when it comes to total crime, it does not have the worst problem, the less per capita. that distinction goes to the neighbor city of emeryville. >>> coming up we look into a story that struck a real nerve with bay area riders. as roadways flooded and schools shut down during storm-geddon, did uber fail to hold its definition of emergency and take advantage of its riders? >>> and plus goi

is a lesson data that said in official publications. drowsy stimates for driving, those sstatistics are based on data is collected and presently are available to 2012. the most recently from 2012 shows that 1.6% of drivers involved in fatal crashes in were believed to be drowsy or fatigued. >> what would you say is a the drowsy nd for driving data? >> in the official statistics there doesn't seem to be much if anything trend, the proportion of the official statistics may have decreased slightly in recent years. however, as i talked about in such a entation, that is vast underestimate that i would published in data those sources for evidence of a trend. i conducted that from 1999 to a 5000 with a sample of crashes each year, in the day that there was no evidence of increasing or decreasing trend in the proportion of crashes that involved drowsiness. in recent years, the raw number injuries and d deaths, nationwide, each year has decreased and that is a good thing. there r, i do not think is any evidence that drowsiness is increasing or d

teft, you cited 2009 data as nhtsa's last official data and i'm wondering if you had a chance to look at anything that might be more recent and if so could comment briefly what it shows? >> right. the 2009, the data from 2009 are the last data that have been cited in official publications by the national highway traffic safety administration as their estimates of the prevalence of drowsy driving. however those data are, those statistics are based on data that are collected annually and presently available through 2012 and the most recent data which i have looked at from 2012 show that 1.6% of drivers involved in fatal crashes in 2012 were believed to be drowsy or fatigued at the time of the crash. >> what would you say is the general trend for the drowsy driving data? let me back up. is there a general trend and if so, which direction is it heading? >> well, in the, in the official statistics there doesn't seem to be much of a general trend. if anything the proportion in the statistics may have decreased slightly in several years. as i talked about in my presentation i believe those a

some of the data we reviewed, i won't go into detail. the first questions is kad for services and a map put in the data set that the work can review or discuss. this was some information that maybe there wasn't information that we could potentially map, but it was something that we wanted to talk about such as the influx of folks to the city during the daytime that's a concern to the p.d. but not necessarily mappable. we did follow all the legislation. if it was available it was reviewed and discussed and data important to the p.d. but not necessarily listed in the legislation such as the location for shootings in addition to homicide. it reviewed each one of those pieces. the selection product brought us over 50 suggestions in those interviews and we created the map for the working group. they reviewed those maps brought them down to nine and within the next month made refinements to and reviewed all the data in the discussion slide that they brought down to four maps that they thought would be workable and then sent to the steering com

a big data and companies, data is now your business.n jewel, it is everything i have got. i may have everything else about my data is increasingly critical to me. choose and why give it to is important. data choosing who i given to is important. about the fact that you should hand it over. why should i keep it in-house was more why not keep in the house? i know who has connections. why don't i do that? >> are you protecting or do you feel comfortable being more efficient than what it would be with a different cloud provider? -- theare deploying cloud is evolving or creating awareness but at the end of the day, the traditional i.t. has stress and people are losing laptops with sensitive information. we eliminate the need of that whole type of technology. people are starting to look. all ofide you access to the security standards, key security things you need in order to deploy. we give you all the measures to evaluate us. probably do it within your own premise. >> ok, stay with us. london traffic, you joined us a bit delayed. we will hol

but they again have access to this data. so to say you fix it, part of me says wait a minute the problem here is my heels of a comprehensive and then you don't even have a handle on it because you have this data. >> that is a great question. we have a chapter. this us back one of the chapters was used. i think that we would all agree that we need new legislation. i think as you all know congress hasn't been at the forefront over the last few years but that is their constitutional role. but we need new legislation and the world i come from a very much agree there was overclassification but now in the new world and it's a valley in seattle but it's scaring me the amount that we can do by putting together an rsi mounds of data and then with those enormous amounts of data find a slight anomalies which we then can exploit. the example that i will give you sophisticated adversaries use mathematical numbers to orchestrate their attacks so that it will only take place over an extended period of time for times that if you saw these f

we use data research and expert analysis to separate the truth from the bunk. let's begin this evening with a pocketbook issue that is impact impactingimpact ing virtually all americans, which is gas prices. gas prices have been plummeting for months, but the oil industry is warning california drivers not to get too comfortable. they say prices are going to soar in 2015 because of a new law. california drivers are accustomed to high prices at the gas pump, but the oil lobby is sounding the alarm claiming they could skyrocket in 2015. is this true? >> this is very much an extreme and very unlikely scenario of the gas prices would go up anywhere near that much. >> reporter: he is an energy policy expert at stanford university who believes gas prices will go up modestly in 2015 thanks to california's landmark cap and trade program. you see so far the program is taxing producers we commonly think of as polluting the air, energy companies like pgne and industrial plants. but the largest source of greenhouse pollution comes from the transportation industry and kickin

condition and arrest and did not collect traffic stop data. another case involved conduct self inflict iv credit, when citing for a traffic violation he sent her four personal text message which she did not respond to and also approached her when she was in court to contest the citation. the officer was suspended. in the final case an officer failed to write an incident report when the crime was reported by a taxi driver and that concludes my report. >> are there any questions for director hicks? >> congratulations in terms of hosting the l.a.p.d. and showing the mediation program that won all of these awards which was a great opportunity and if good for you in the office. congratulations. >> thank you very much. the l.a.p.d.'s mediation program is very different than the occ. the mediation program is run through their city attorney's office, but it is achieving success and they do have a limited jurisdiction. they are only limiting, they are limiting their mediations to only racial profiling allegations. i'm sure that they will seek to expand

my data comes cheap.but as we start having these more connected devices we start collecting more and more data. some manufacturers will throw that data in the cloud, and data in aggregate always has more value than data dispersed to the endpoint because if you are collecting data to disperse to the endpoint it is labor intensive so it has to be more valuable to make it worth the effort. and that is what i think that we will see play out. what information that could be monetized will be aggregated and has become the new targets. >> what do we think in 2020, you know, cybercrime is all over the world, the bulk of practitioners seem to be in eastern europe, and there are good reasons for that. i would like to throw out to this group, is that work still going to be? it is in eastern europe for lots of reasons because of legal reasons that they can operate freely. it is also their because their are smart people in that area of the world who are trained in some of the best universities in the world and did not

we get a lot of data out today.? >> i'm watching right now, i think home prices are okay. i think gdp is going to be even bigger. but i think the key story here is we have an ongoing accelerating recovery. we at commonwealth i think do expect markets to go higher in the short-term. >>> and the s&p 500 settling at a record high for the 50th time this year. if the price of oil continues to drop, could that derail this market rally? >> if you actually look at the numbers, there is a slight negative effect from i'll price drop. you see that in reduced business investment and potentially reduced employment. but the cost of effects outweigh that by about six to one. this is a huge benefit not just for the u.s., but for the woerth as a whole. i think that's going to accelerate the economic activity. >> the feds, we don't want to fight the feds. you think rates are going to head higher this year, i'm assuming. do we reposition, just ride with it? >> i think you need to be a little defensive in your fixed income allocation.

my assumption is staff did not generate this mix along the data that supports it and the data that is not available to prove this. several numbers regarding risk, this change that was proposed to us back in june that started in february was trying to reduce the total risk the fund the long risk that we have. but at the same time trying to keep our eye on the distributions. it's not simply about reducing risk but trying to achieve a total return that is fair to the members because we are all the active members are contributing more and fairness to those paying the majority of the contribution. hedge funds, this idea came along because it was designed to reduce risk. i believe i said back in may or june we wanted to educate and thanks to the board and the staff. the numbers i'm look ing at show, i don't want to say asset class because hedge funds are a collection of strategy. it actually less risky. it is less risky. there have been a lot of other erroneous data out there. i'm not ignoring the fee issue. it's more consistent return so the city can plan it's budget and not just to guaran

is collected analysis the data i think we're going to see a lot more progress and strong commitment to work with the police department to get the data we'll see some very good progress made on the education accident john mentioned the citywide education strategy we started with the sf pledge and have the larger vehicle i want to the thank paul from the ecology team that focused on helping us there's much more work to do but we can break ground quickly by disseminate nationally it to groups and trying to figure out how to institutionalize vision zero and the funding it is difficult to fund those programs there is a mechanism but the applicable accomplishment we've seen working with the traffic and safety a strong commitment to figure out a strategy we'll be exploring that in the next year an evaluation monitoring that is to answer the questions how effective are the policies we've been fortunate that we've moved some done bureaucratic things how to fund a position with the mta and public health we've worked with public health department on the work order to help with the team do the

are we putting it in the field or the data and at a very basic level that is the explanation and the reason that something gets mapped to the wrong field is the data base came up in pts, which is basically the scheme is basically the structure of how all of the data is stored in multiple tables and then there are relationships between the tables that is a system that has been maintained and enhanced over its lifetime, 14 or 15 years. all of the folks that have performed the maintenance over that period are not all with dbi and so they have retired and moved on to other jobs and so, even understanding the pts data structure is some what of a process of discovery, there is a lot of expertise on the team, but there are some aspects to be able to move everything to a new platform. the as i talked about the specialized user training which was led by an ainstructor that was completed by the beginning of december and we have refresher training and a couple of rooms set up at dbi where we have got folks from every department, that come in whenever they have schedule the time and just basical

maybe the problem isn't the data. maybe the problem is the level of authentication we give data, the possession updated when we shouldn't. maybe the problem is in the transaction. >> good enough spent if you look at the date of breach report to reproduce, laughter we looked at just over 63,000 security incidents, just over 1300 data breaches from a 50 participants in 95 countries. and i think when you look at it consistently year over year, two factors become a major issue. if most folks, i think 90 plus percent of all things we see would be resolved by leveraging. >> which is getting easier with the new communications. much easier. >> so we've now reached 2020, and i want to talk about what cybercrime looks like in 2020. so dino, what does it take to be a crop -- a cybercrime and 2020 what do we need a computer science degree to? >> either that or any accredited cybercrime university. but one thing that's different shaded cybercrime from a lot of other fields is a lot of the skills can be learned in the underground a