structured versus unstructured data, tagging data, i leave that to my data science friends to figure out how to do that. the important thing is the integration of information that we have to build new knowledge. the buzzword after 9/11 was connecting the dots. we are not only connecting dots anymore, we are connecting the needles and haystacks under needles and haystacks. and needles and haystacks to try to get answers. the challenge is moving data, protecting civil rights and civil liberties, not making false allegations and trying to do all that in real time. if you got an idea, let us know when we will try to answer it. it is a daunting challenge, but i would tell you that the u.s. government and our partners across the globe have taken on that challenge in ways that we never thought we could do, in protecting the information, protecting civil rights and civil liberties, and also sharing that information in near real-time. much better than 15 years ago after 9/11. >> we talked a lot about the political security and social aspects of countering terrorism. no one on the panel has sp

experts in bringing unstructured data into structured data and only when the data is structured can we do any meaningful analysis. >> what kinds of things are you looking for? or what have you found? >> so it is a little bit premature to tell. we've been looking at 89 fields that the d.a. department -- the san francisco d.a. told us are of interest. we've been putting those in to a structured database. now the analysis is really only about to begin. we've learned -- i'm not a psychologist, but to show -- i think the ultimate goal of this project is to create a screening tool for prosecutors where they could potentially look at police reports and identify police reports that might be based on bias. >> when will you have the data available or the analysis available? >> so i think we are looking at a multi-year research project. >> smuure. sure. >> and that will have to involve not only machine learning experts but also criminologists, psychologists to really understand what the patterns we are detecting, what they actually mean. then at the next step, how should those patterns inform pro

structured versus unstructured data, tagging data, i leave that to my data science friends to figure out how to do that. the important thing is the integration of information that we have to build new knowledge. the buzzword after 9/11 was connecting the dots. we are not only connecting dots anymore, we are connecting the needles and haystacks under needles and haystacks. and needles and haystacks to try to get answers. the challenge is moving data, protecting civil rights and civil liberties, not making false allegations and trying to do all that in real time. if you got an idea, let us know when we will try to answer it. it is a daunting challenge, but i would tell you that the u.s. government and our partners across the globe have taken on that challenge in ways that we never thought we could do, in protecting the information, protecting civil rights and civil liberties, and also sharing that information in near real-time. much better than 15 years ago after 9/11. >> we talked a lot about the political security and social aspects of countering terrorism. no one on the panel has sp