Final Presentation – Digital Society for All

So this is the report
from group, and I am doing the presentation. But this is contributions from
a lot of people in the group. And all of you who are here,
please feel free to jump in, if I'm saying something wrong or
missing something. So, what we came up
with was a proposal to build a Nation-wide
complaint management system. So, if you have been living
in India for a long time, you will find that there lots
of things to complain about. And we think one big problem
is that you do not know where, or how to submit a complaint. And even if you do
know of a particular website comment certainly take
a comment website where you can submit that complaint. The form is very conversive, you
will have to really fill in many fields, and you don't know what
to expect at the end of it. And that is from
the citizen site, and also from the government's
department site. This is get this
barrage of complaints, many of them are in forms you
cannot even know what is it that that person is
complaining about. I have actually personally
looked at the complaints submitted on Maharashtra, you
know Mumbai website, and there's such so many spelling mistakes,
grammatical errors, I don't know what exactly is the complaint,
what is being talked about. So there are many ways in which
we can help both parties, by creating this one
unified nationwide company management system. So we are thinking of it like
a stack overflow for complaints. Or a quota for complaints. Or maybe somewhere in between
Wikipedia for complaints. So, once we have such
a complaint, once we have this interlaced platform, it will
accept inputs in the form of, maybe, you can call
a 311 kind of number, so you'd be getting audio input. And I think the primary means
of complaint submission will be to mobiles. You take a picture and, along with your GPS coordinate,
submitted. Or you can just sort
of text something. Maybe we have handles with
Twitter and Facebook. So that you don't necessarily
have to install our app to get a complaint. To submit a complaint. And then we'll be, of course, connecting to existing systems
for complaint management. So, we should have hooks for
that. And once you cleared this
connection, this network of submitters of complaints and
solution providers. You can do several things,
I mean you have many of the things that I
presented in my Inventor Talk. Plus, there's so many other
things which I did not present, like, you could dig out
other relevant solutions to previous history of complaints. Very often what happens
is people say, okay, your complaint has been fixed. But actually, it resurfaces. Somebody said,
we fixed this pothole but within one month again
there is a pothole. So can you resurrect
old complaints? And then from sort of say,
election candidate's point of view,
you could provide analytics over the history of complaint
resolution in an award. And it definitely
has relevance and support from people trying
to build smart cities, and now of course,
there will be many hurdles. Question is, how would you incentivize the service
providers. So, people had some nice
ideas and we had made a list. We could maybe use public
shaming, or with lawsuits. They are the friendly startups
which specialize, and which will help you do lawsuits. And the thing is, can we involve
this large layer of private agencies to help mitigate
some of these complaints. Maybe for a fee somebody will. Okay, you really cannot bear
the stink coming from the drain outside your door, and
you're willing to maybe, be a private agency to achieve your
expected level of sanitation. If you can have private
security guards, provide private police officers, why can't you have private,
such as sanitation agencies. Now from AI point of view, there
are many research challenges, of course understanding audio,
image. NLP and the text complains
is not going to be easy. So we have to take the easy
to input complaint format. And transform it into a form
that a government agency might meant. So people such as Ted, can we
have Task-focused chatbots. Like when I was looking at
Mumbai municipality complaints, many of the complaints,
you just could not find, figure out what they're
complaining about. So a chatbot can cross question
to get more details as when required. But how it's been to and to see what we talked
about a lot of time. Somehow, I do believe that for
a complain manual system it's going to fly
only when it is very fast, and you can just go it on the go. So, and then there were nice
ideas about how we could use crowd sourcing, to both help
with understanding of the input, and also to sort of
to designate expert. So somebody might to
take more some example. Somebody might declare himself,
or herself to be the expert on water related issue,
in a particular war in Delhi. And then can we have some kind
of a Wikipedia like model, for controlling that expertise. And then, there were ideas
about actionable news. For example, in Mumbai, there was this well,
just atrocious stampede. And then, I changed by city
which is this complain management portal which
is I didn't existed. They actually took that news,
and they were conducting survey among my city
registered users to say what could have been done,
what representation can we make. Do you have any ideas on what
could have been done to not have such accidents
again in future. And of course, multi linguality. And another very nice project
that is kind of a project on its own, is how do you get
verification of whether a complaint has been
dissolved or not? And this can actually be
a much bigger project, so suppose if someone says
pothole has been fixed, how do you involve citizens to validate whether that
actually has been done or not? And I believe Microsoft all
ready has some projects of that nature. So that team can work
together with this effort, to kind of add validation with
the citizen supported crowd sourced to the resolution
of complaints. And now of course the difficult
is how to make it happen. So [LAUGH] basically we
taught what we can try to do. I mean initially,
it has to be a research project. Because there are many
research challenges. So maybe it can be like, there was interest from IT
current pool, and from MSR, and definitely IT Bombay is
interested, and IT Delhi, so we could create a GitHub
project with maybe, perhaps, the right software architecture,
so that different teams can
contribute to different pieces. We all ready have sort of some
kind of a system around winter. So we can just make
it extensible and easy to work on
by lots of teams. And Microsoft,
we are hoping will help us in choosing the right
development platform, which will allow for large
scale sharing and deployment. And then we will, and of course, the next big challenge is
how to find developers? We do find students once in
a while who are good in coding. But here, what you need, actually I am familiar with the
old style software development. We used to have people who
were specifically hired, they're like master developers,
people who did BST, one of those staff would be
the post risk project staff. They would be responsible. They are not grad students. They are not faculty. Just know the system really
well they maintain, and they do maybe 50% of
the coding of that system. So, I don't know whether
we can find such a person, or well, first,
more than find actually, I think just finding
such people is difficult. And particularly in
a university setting. There is not much place in
university growth path for people who are neither
students nor faculty. In other words,
what am I doing here? So that is the problem with
our finding developers. And we are hoping that
the researchers involvement will be done, when your building
an initial good prototype, and then we are looking for
more more partners, enablers, like she does, to kind of reach
out to government agencies and selling it to
government departments, selling it to people and
making it happen. This is the main thing
that we discussed, but there was anything interesting
project, I don't know if somebody is interested in,
you know just sort of elaborating more on
the cities and even. Evaluation of civic projects. So this is something, actually I don't maybe don't
want to talk about it much, because that was mostly what
I wanted to talk about. Yeah thanks. >> [APPLAUSE]
>> But is there anything
anyone wants to add? >> I think it's easy to
build an application. It's perhaps even possible to
get departments to adopt it. But getting people to adopt it. Getting consumers, is
basically a very hard problem. >> But you know,
I change my city. At least, I don't know what
is the fraction of people who use it, but in terms of-
>> Let's take a poll in this room. I mean,
how many people from Bangalore. Music, for example. >> I don't know, how many
have even heard about it? >> So I have a feeling that
marketing it to people is where the real challenge is. I mean there are a lot of apps,
for example the government of India,
the central government, has been promoting the which
does the same thing. And then, I know that the
Connota state government has at least two independent portals
that are developed by different departments. So I feel the adoption and
preventing fragmention and all of that, that is going
to be the real challenge. And then we can get to the
interesting problems behind it. >> Yes, yeah. >> Yeah, just to add it
actually government of Connota launched [INAUDIBLE]
a few years back called >> [FOREIGN]? >> [FOREIGN] stands for actually basically the main
object of the application is to guarantee delivery of
services to its citizens. So that is being
actually like exactly what you portrayed [INAUDIBLE]
in terms of direction, right? That is kind of actually like
interface which should be directly connected to the
comment and as well as system, and the entire backing the
mechanism in terms of addressing it to the right official in
that particular department. And making sure that how
that service is rained out gives the logical end to that. And also it involves
the citizens' feedback for the closure of that
particular activity. It can be like issuing of
the birth certificate or actually services or
property tax. Any of those close to 300
different services which a typical government
office to the citizen. So that is there in place, and
also like effectiveness is they're getting about close to
10,000 to 15,000 hits every day. And they are actually monitoring
it in terms of seeing what is a [INAUDIBLE] associated with
that particular service. And which are delivering
within their SLS, and which other departments
are failing to do that so that they can bring in
the course collection. So maybe we can look at
actually integrating that. That would be a good
platform for taking forward. [INAUDIBLE]
>> For winter, we didn't have any
data to start with. We basically. [INAUDIBLE]
So we got all their
data [INAUDIBLE] so literally [INAUDIBLE]. >> And also a lot of comments. Comments of experiment with
a lot of [INAUDIBLE] analysis. They're making the [INAUDIBLE]
in terms of which are the main pain points of this one, it can
be pot hole, it can be poor or it can be a rain
problem water logging. So that has been done
on an experimental basis in government of India,
in fact we were involved in that in terms of
developing the dashboard. Especially when they had
the bigger events like, and all those things right which
occur when we get about huge courage in the crowd. So they did actually make
something called real-time monitoring systems where they
got involved in about four or five departments
mainly involved in terms of actually bringing
that kind of thing. So we can approach those
departments taking those frameworks, they'd be more
than happy to take it forward, actually, yeah. >> [INAUDIBLE]
>> I'm from Microsoft. I work with IDC. I've been handling the
government segment business for the last 15 years
within Microsoft, so we are actually more this kind
of from the government, yeah.

1 thought on “Final Presentation – Digital Society for All

Leave a Reply

Your email address will not be published. Required fields are marked *