Influence Mapping – Dead Project?

[Minitrue]

Hello everyone,

We're delighted to discover such an honourable project, but it seems from the website's blog, and from the scarce activity on this forum, that entropy has set in.

Is there anyone out there who is still interested in moving forward with influence mapping?

Please get in touch.

M

[Friedrich Lindenberg]

Hey, 

I think a lot of the people on this list are actively working on influence mapping-related projects, the original project behind the community just ran out of steam (and funding), for example on the amazing blogs and demo sessions coordinated by Paola. 

Maybe as a little update from my end, I’ve been working for the OCCRP (https://occrp.org) on their data search engine: https://data.occrp.org/ - it’s an attempt to make as much data as we can get our hands on visible and accessible to investigative reporters in our network. The underlying technology is open source here: https://github.com/alephdata - We’re hoping to add much more investigation modelling to this, including project management of persons of interests, their relationships, time lines etc.

Another strand I’ve been working on is collecting and consolidating international sanctions lists. That project (http://www.opensanctions.org/) is trying to create an open baseline dataset for due diligence-type tasks by collecting up-to-date information about persons of interest from political, business and criminal backgrounds. 

What have other people been up to?

- Friedrich

--
You received this message because you are subscribed to the Google Groups "influence-mapping" group.
To unsubscribe from this group and stop receiving emails from it, send an email to influencemappi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Wills, Tom]

Hi Friedrich and all,

On 25 June 2017 at 16:08, Friedrich Lindenberg

<friedrich....@gmail.com> wrote:

> Another strand I’ve been working on is collecting and consolidating
> international sanctions lists. That project (http://www.opensanctions.org/)
> is trying to create an open baseline dataset for due diligence-type tasks by
> collecting up-to-date information about persons of interest from political,
> business and criminal backgrounds.

OpenSanctions looks very cool - nice to see it has come together so quickly.

An idea for a further source of persons of interest: In the past I've
used SPARQL queries against DBpedia to extract names of interest from
Wikipedia.

For example, you can potentially devise a query that returns all
national-level football players. Thanks to the infoboxes on Wikipedia,
there is a certain amount of structured data available, although the
quality is patchy. Perhaps too patchy to include in something like
OpenSanctions, I admit.

I might be able to give it a try at some point, but thought I'd share
in case anyone else wants to give it a go.

Tom
--
Tom Wills
Data Journalism Editor
The Times & The Sunday Times
Direct line: +44 (0)20 7782 5776
tom....@thetimes.co.uk

--



--

"Please consider the environment before printing this e-mail"

Newsworks - bringing advertisers and newsbrands together

www.newsworks.org.uk

This e-mail and any attachments are confidential, may be

legally privileged and are the property of News Corp UK & Ireland

Limited on whose systems they were generated. News Corp UK

& Ireland Limited is the holding company for the News UK group,

is registered in England & Wales under number 81701, has its

registered office at 1 London Bridge Street, London, SE1 9GF and

is registered with VAT number GB 243 8054 69. If you have received

this e-mail in error, please notify the sender immediately and do not

use, distribute, store or copy it in any way. Statements or opinions in

this e-mail or any attachment are those of the author and are not
necessarily agreed or authorised by News Corp UK & Ireland Limited

or any member of its group. News Corp UK & Ireland Limited may

monitor outgoing or incoming emails as permitted by law. It accepts
no liability for viruses introduced by this e-mail or attachments.

News Corp UK & Ireland Limited and its titles are committed to abiding by
IPSO's regulations and the Editors' Code of Practice that IPSO enforces.

[Friedrich Lindenberg]

Hey, 

On 26. Jun 2017, at 13:34, Wills, Tom <tom....@thetimes.co.uk> wrote:

Hi Friedrich and all,

On 25 June 2017 at 16:08, Friedrich Lindenberg
<friedrich....@gmail.com> wrote:

Another strand I’ve been working on is collecting and consolidating
international sanctions lists. That project (http://www.opensanctions.org/)
is trying to create an open baseline dataset for due diligence-type tasks by
collecting up-to-date information about persons of interest from political,
business and criminal backgrounds.

OpenSanctions looks very cool - nice to see it has come together so quickly.

An idea for a further source of persons of interest: In the past I've
used SPARQL queries against DBpedia to extract names of interest from
Wikipedia.

For example, you can potentially devise a query that returns all
national-level football players. Thanks to the infoboxes on Wikipedia,
there is a certain amount of structured data available, although the
quality is patchy. Perhaps too patchy to include in something like
OpenSanctions, I admit.

I might be able to give it a try at some point, but thought I'd share
in case anyone else wants to give it a go.

We’ve played with that a good bit at OCCRP and it’s actually proven to be quite useful. Basically going off WP category pages, in different language editions to bootstrap a PEP list. It worked really well for businesspeople in some countries, but sometimes goes completely off the rails and then starts adding 13th century knights to your database. But I’d love to include it in OpenSanctions, if you feel keen to do a crawler that would be fantastic. 

Here’s a super-old version of my code that someone cloned ;) https://github.com/backgroundcheck/wikipeps 

- Friedrich

[Kevin Connor]

Hi all,

Friedrich is right, a lot of people are involved in influence mapping projects still but the project ran out of steam/funding. An update on my end – I am still working on LittleSis and focusing a lot of time and energy on various power mapping projects, mostly in the US. We put up this page for an improved Oligrapher, our open source mapping tool, which you can also install and use with non-LittleSis data sets (this was made possible by the Influence Mapping project!). Here is the page: littlesis.org/oligrapher (complete with enthusiastic instructional videos)

We also have been doing some projects with activist-researchers interested in mapping power around Trump. This has included a series of webinars and a toolkit. We've been training/working with groups of activists and journalists in different cities who are doing power mapping/research intended to support challenges to Trump and the corporate/billionaire networks clamoring to, among other things, kick people off healthcare so they can get tax breaks.

We've been doing lots on the tech side to make the LittleSis software more stable/reliable/usable, but I will spare you those updates. A more exciting one is that we developed a chrome extension, which we just put up, that facilitates in-page editing of LittleSis.

Kevin

On Mon, Jun 26, 2017 at 8:23 AM, Friedrich Lindenberg <friedrich....@gmail.com> wrote:

Hey, 

On 26. Jun 2017, at 13:34, Wills, Tom <tom....@thetimes.co.uk> wrote:

Hi Friedrich and all,

On 25 June 2017 at 16:08, Friedrich Lindenberg

--
You received this message because you are subscribed to the Google Groups "influence-mapping" group.

To unsubscribe from this group and stop receiving emails from it, send an email to influencemapping+unsubscribe@googlegroups.com.

[stef]

ohi,

some questions below

On Mon, Jun 26, 2017 at 11:10:53AM -0400, Kevin Connor wrote:
> Friedrich is right, a lot of people are involved in influence mapping
> projects still but the project ran out of steam/funding.

what project? what funding? i never knew about either.

anyway, i'm still also doing related stuff.

i also wonder what is the plan with this chinese propaganda that was posted
lately on the list?

cheers,s

[Friedrich Lindenberg]

Hey stef,

On 26. Jun 2017, at 17:16, stef <s...@ctrlc.hu> wrote:

ohi,

some questions below

On Mon, Jun 26, 2017 at 11:10:53AM -0400, Kevin Connor wrote:

Friedrich is right, a lot of people are involved in influence mapping
projects still but the project ran out of steam/funding.

what project? what funding? i never knew about either.

http://www.influencemapping.org/ :)

anyway, i'm still also doing related stuff.

Cool, what?

i also wonder what is the plan with this chinese propaganda that was posted
lately on the list?

I’ve banned a few users. Sorry, but I didn’t see any added value. 

- Friedrich

[stef]

On Mon, Jun 26, 2017 at 05:18:26PM +0200, Friedrich Lindenberg wrote:
> Cool, what?

mostly parltrack and commission lobby/meeting register stuff

[Ben Parker]

Dear Kevin

Thanks for the update on the good stuff over at LittleSis!

I and others at OpenCorporates have been accumulating a list of the companies associated with the various Trumps and Kushner, past and present, which can be filtered by country and US state and downloaded via API etc. It's at 1,055 entries as of today for Trump himself. You might want to add to your toolkit:

https://opencorporates.com/corporate_groupings/Donald+Trump

https://opencorporates.com/corporate_groupings/Jared+Kushner

https://opencorporates.com/corporate_groupings/Ivanka+Trump

https://opencorporates.com/corporate_groupings/Donald+Trump+Jr-

https://opencorporates.com/corporate_groupings/Eric+Trump

Best wishes

Ben

On Mon, Jun 26, 2017 at 5:10 PM, Kevin Connor <kpco...@gmail.com> wrote:

Hi all,

Friedrich is right, a lot of people are involved in influence mapping projects still but the project ran out of steam/funding. An update on my end – I am still working on LittleSis and focusing a lot of time and energy on various power mapping projects, mostly in the US. We put up this page for an improved Oligrapher, our open source mapping tool, which you can also install and use with non-LittleSis data sets (this was made possible by the Influence Mapping project!). Here is the page: littlesis.org/oligrapher (complete with enthusiastic instructional videos)

We also have been doing some projects with activist-researchers interested in mapping power around Trump. This has included a series of webinars and a toolkit. We've been training/working with groups of activists and journalists in different cities who are doing power mapping/research intended to support challenges to Trump and the corporate/billionaire networks clamoring to, among other things, kick people off healthcare so they can get tax breaks.

We've been doing lots on the tech side to make the LittleSis software more stable/reliable/usable, but I will spare you those updates. A more exciting one is that we developed a chrome extension, which we just put up, that facilitates in-page editing of LittleSis.

Kevin

On Mon, Jun 26, 2017 at 8:23 AM, Friedrich Lindenberg <friedrich.lindenberg@gmail.com> wrote:

Hey, 

On 26. Jun 2017, at 13:34, Wills, Tom <tom....@thetimes.co.uk> wrote:

Hi Friedrich and all,

On 25 June 2017 at 16:08, Friedrich Lindenberg

[Kevin Connor]

Thanks Ben! I already saw it and found it super helpful, but yes we'll have to add. Thanks for doing that.

[Minitrue]

Delighted to hear people are still working on these projects. Is there any way of combining all these different databases into one searchable master database? I'm thinking type in a name and you can visualizing all the connections of that node with other power nodes.

[James McKinney]

Hi M,

That's the purpose of the proof-of-concept "Who's Got Dirt?" API: http://www.influencemapping.org/work/whosgotdirt/

Best,

James

[Khairil Yusof]

On Sun, 2017-06-25 at 17:08 +0200, Friedrich Lindenberg wrote:

Hi Friedrich,

We're working on this too with Popit (Popolo-spec), and we've added
Relations class to it already, to be able to also model relationships
flexibily at least all relationships as usually defined by PEPs
definitions.

I was about to shoot an email off to ICIJ (Asia) to update them on our
work, and had similar ideas on having single global databases to search
with (great initial work by OCCRP btw, love the invetigative journalism
dashboard!)

We're actively working, but like others limited funding and resources
means, we can't be as active in outreach as we would like to.

I have draft of modeling and categorising PEPs in Popolo-spec for
Malaysia that's just sitting collecting dust, as I haven't had time to
work and implement some test cases on it yet.

> Maybe as a little update from my end, I’ve been working for the OCCRP
> (https://occrp.org) on their data search

> engine: https://data.occrp.org/g/ - it’s an attempt to make as much

> data as we can get our hands on visible and accessible to
> investigative reporters in our network. The underlying technology is

> open source here: https://github.com/alephdata&nbsp;- We’re hoping to

> add much more investigation modelling to this, including project
> management of persons of interests, their relationships, time lines
> etc.

https://github.com/Sinar/popit_ng

[Friedrich Lindenberg]

Hey Khairil,

thanks so much for getting in touch about this. I’d love to see how you’re modelling PEPs, since we’re also doing a fair bit of that, and I’ve been thinking about whether it would make sense to create a sister project to EveryPolitician which would attempt to aggregate additional details, such as relationships, family and non-politicians from national-level projects like yours (or the amazing thing people did in Ukraine - did you see https://pep.org.ua/en/?).

For what it’s worth, you might also be interested in the ontology we’ve been using to model OCCRP’s data. It’s probably more of a cousin to Popolo than a child, the main goal is to be pragmatic and allow us to describe the data we have: https://alephdata.github.io/followthemoney/ (to be super clear: this is not an attempt to create a data standard. This is a format for application data. Please don’t elevate to a data standard - popolo is the truth :)

Cheers, 

- Friedrich 

[Khairil Yusof]

This looks really good.

https://alephdata.github.io/followthemoney/

This was something of a next step for us, as in first we want a robust
way to to build and track PEPs, over time (years maybe) and with bits
and pieces of info from different sources.

Then to be able to use that info for something like alephdata follow
the money.

So for Popit 2, we differed from EveryPolitian in that we make the
assumption that there is not going to be a nice complete list, and that
you will be working with incomplete set of best available data at the
moment. Popolo spec database keeps that structure for us, while the
pieces are missing.

eg. PEP A is known to be Director on Company B. But we might not know
his associates in Company B until some other leak or data release.

We also needed to track what position or which PEP was in that position
for abuse of power for personal gain at a particular date in time.

We rely on Popolo Posts heavily for that.

We planned to better model PEPs in Popolo by using the role in Posts
and Organization category as a form of categorization.

A person would be a PEP when they held a role as Parliamentarian, but
they might also be a PEP at a different time when they held post with
role of Director of Government Agency.

We use Popit this way, not as a lists, but tracking a person holding
positions and their relations across time (like Cargografias)

As my other email, the only limitation we faced with Popolo-spec to
model PEPs was lack of Relations, for which we implemented a schema
with feedback/input from James. This allowed us to include family,
business associates etc. who did not hold any posts.

I forgot one more thing we added to our database feature was per field
citations. So that every bit of piece of info has provenance. As some
of these PEPs especially business associates, special officers, private
secreatries are shady characters and it's not easy to get their
details.

On Mon, 2018-04-02 at 12:12 +0200, Friedrich Lindenberg wrote:
> Hey Khairil,
>
> thanks so much for getting in touch about this. I’d love to see how
> you’re modelling PEPs, since we’re also doing a fair bit of that, and
> I’ve been thinking about whether it would make sense to create a
> sister project to EveryPolitician which would attempt to aggregate
> additional details, such as relationships, family and non-politicians
> from national-level projects like yours (or the amazing thing people
> did in Ukraine - did you see https://pep.org.ua/en/?).
>
> For what it’s worth, you might also be interested in the ontology
> we’ve been using to model OCCRP’s data. It’s probably more of a
> cousin to Popolo than a child, the main goal is to be pragmatic and
> allow us to describe the data we have: https://alephdata.github.io/fo

> llowthemoney/ (to be super clear: this is not an attempt to create a

> data standard. This is a format for application data. Please don’t
> elevate to a data standard - popolo is the truth :)
>
> Cheers,
>
> - Friedrich
>
>

> > On 30. Mar 2018, at 06:31, Khairil Yusof <khairil.yusof@sinarprojec

[Friedrich Lindenberg]

On 2. Apr 2018, at 15:41, Khairil Yusof <khairi...@sinarproject.org> wrote:

So for Popit 2, we differed from EveryPolitian in that we make the
assumption that there is not going to be a nice complete list, and that
you will be working with incomplete set of best available data at the
moment. Popolo spec database keeps that structure for us, while the
pieces are missing.

eg. PEP A is known to be Director on Company B. But we might not know
his associates in Company B until some other leak or data release.

We also needed to track what position or which PEP was in that position
for abuse of power for personal gain at a particular date in time.

The multiple sources bit is where I’ve been struggling. So with the current alephdata our focus has been to represent each source data item as well as we can, and the next step will be to make synthetic combined entities about PEPs that merge the properties and relations from multiple source entities into a “dossier”. 

A person would be a PEP when they held a role as Parliamentarian, but
they might also be a PEP at a different time when they held post with
role of Director of Government Agency.

We use Popit this way, not as a lists, but tracking a person holding
positions and their relations across time (like Cargografias)

Do you have good frontend tooling to represent that? I’m extremely interested in timeline tools that can be used by investigative reporters to build event timelines (whether biographical or commercial transactions).  

As my other email, the only limitation we faced with Popolo-spec to
model PEPs was lack of Relations, for which we implemented a schema
with feedback/input from James. This allowed us to include family,
business associates etc. who did not hold any posts.

I forgot one more thing we added to our database feature was per field
citations. So that every bit of piece of info has provenance. As some
of these PEPs especially business associates, special officers, private
secreatries are shady characters and it's not easy to get their
details.

To me, that’s the most impressive aspect of WhoWasInCommand (also with James’ contribution): https://securityforcemonitor.org/whowasincommand/ - they seem to be sourcing each data point to an article, and can tell how many pieces of corroboration exist for each of the data shown in the tool. It’s impressive, but I’m a little worried about coming up with good enough data entry tools and incentive systems to keep people doing this.

I’m super interested in how you have modelled family and other informal relationships, is there anywhere I could see your model for that? 

- Friedrich 

[Khairil Yusof]

On Mon, 2018-04-02 at 18:17 +0200, Friedrich Lindenberg wrote:

<snip>

> The multiple sources bit is where I’ve been struggling. So with the
> current alephdata our focus has been to represent each source data
> item as well as we can, and the next step will be to make synthetic
> combined entities about PEPs that merge the properties and relations
> from multiple source entities into a “dossier”.

This is why we use Popolo and OCDS for the "structure" of the different
pieces. But we're only using pieces, to only build PEP (popolo) info
and later the same approach for Contracts (OCDS)

I think aleph is on another level above what we're working on. ie.
you're linking PEPs to other data and sources, for complete dossier.

We're just treating the other parts as "sources" to build a complete
picture of our PEP or Contract, but you're organizing the sources
themselves into a complete dossier.

In the same wasy for a Contract. We don't know all the details, so
we're sourcing different info into OCDS structure to build a profile of
one complete contract.

Only then are we even thinking about linking it with other information.
I think we're mostly happy with what we want to do for PEPs right now,
just need a bit more testing on Relations.

I think where we would fit is as a database source solely for PEPs to
be join them up with aleph to other sources, since our PEP model has
history of posts.

> Do you have good frontend tooling to represent that? I’m extremely
> interested in timeline tools that can be used by investigative
> reporters to build event timelines (whether biographical or
> commercial transactions).

Sadly no. Right now we don't have enough Post data (start and end Post
dates) for may of our PEPs, but I had enough for test and POC:

Jupyter Notebook code example on searching for specific PEPs on
specific dates of determining 1MDB PEP from US DOJ filings:
https://sinarproject.org/en/updates/uncovering-1mdb-with-popit-open-dat
a

> > As my other email, the only limitation we faced with Popolo-spec to
> > model PEPs was lack of Relations, for which we implemented a schema
> > with feedback/input from James. This allowed us to include family,
> > business associates etc. who did not hold any posts.
> >
> > I forgot one more thing we added to our database feature was per
> > field
> > citations. So that every bit of piece of info has provenance. As
> > some
> > of these PEPs especially business associates, special officers,
> > private
> > secreatries are shady characters and it's not easy to get their
> > details.
>
> To me, that’s the most impressive aspect of WhoWasInCommand (also
> with James’ contribution): https://securityforcemonitor.org/whowasinc

> ommand/d/ - they seem to be sourcing each data point to an article,