Categories
Digital Newsgathering Journalism Social Media Verification

Making video material verifiable – identifiable features

Reading Journalism and a world in transition: Wadah Khanfar’s James Cameron memorial lecture I thought I’d pick out this passage which relates to how Syrian activists have helped make material easier to verify:

The repetitive doubting, for instance by the Syrian authorities, of what al-Jazeera transmits of news and pictures received via Facebook, Twitter or YouTube and the accusations levelled against us that we fabricate such material, all of this inspired activists to come up with new ideas including a better documentation and authentication of demonstrations.

This they do by means of including in their footage the names of streets or images of landmarks or well-known public squares in the various cities; as well as including the date of each clip using a newspaper front page where the date of issue is clearly printed.

This is definitely something I’ve noticed in recent months. Below are some thoughts on how to extract this information usefully and the sort of things I look out for in a non-exhaustive, non-exclusive list.

The physical environment as source of verification

Physical objects that often people take for granted have become a very useful set of keys in my verification toolbox.

Landmarks and squares are pretty easy to pin down – tourists tend to take photos of them which we can use as reference. Plus they are relatively easy to spot on satellite imagery and to match up to wide shots in video clips.

When it comes to landmarks in the arab world, mosques are a favourite of mine as minarets tend to be distinctive and so easier to identify. These are the sort of buildings for which there tend to be photos I can use as a reference.

They’re only beaten by historically significant buildings or monuments for ease of identifying features I’m glad to see. Rarely do these go un-photographed.

Statues are absolutely brilliant on the grounds of their uniqueness – even if they are a common typology, their setting is usually unique.

Bridges are also helpful landmarks as anything filmed from a bridge tends to mean that there’s a clear line of sight onto the surroundings.

Street names, I’m less a fan of as it’s not always the case that the local name matches up with the one found on maps of the area.

Getting round the language divide

As a non-Arabic speaker, descriptive data that I can stick through google translate is always helpful. I’m fortunate enough to be able to rely on arabic speakers at work to translate handwritten dates and newspaper dates.

When one isn’t available there is an emergency workaround, which is a bit circuitous, so bear with me…

  • Take the claims being made about location and date as you understand them or have translated from captions
  • Highlight keywords in your translation (date, location) and check the original arabic it relates to
  • Make a note of this arabic on paper and keep this in front of you
  • Re-watch video clip looking for a visual match with the date and location information you have in front of you
  • Find someone to give a second set of eyes to the match and confirm the matched information

This method is far from foolproof, but in emergency it may be enough of a level of verification to get you moving in the right direction. But given its slightly blurry results, I don’t use it if I can avoid it at all.

Note of caution

There have been examples where the dates, times and locations claimed on videos haven’t matched up, so it’s useful to do a longitudinal search for videos in the same location to ensure that the clip isn’t a reboot of older material.

This rebooting, in my experience, isn’t deliberate so much as a result of the sometimes unclear path via which material emerges. They don’t get uploaded chronologically or, well, logically at all. It’s as they surface.

But it can be deliberate if someone believes that it corresponds with a reported event and they feel that a clip can be used to represent it. I’ve seen this where a protest was illustrated with video from a similar incident in the same location a week previous.

In that sense it’s always good to have at least two indentifiable features that tally and are consistent.

STAY SAFE

Most of all, stay safe. I say this to everyone I speak to who is filming or photographing events. Even the most ordinary of events can pose a risk – for example tripping over a kerb while filming something across the road.

Small World News have done a nice guide to keeping safe when “in the field” which is worth downloading and reading.

Categories
Digital Newsgathering Journalism Social Media Verification

#bbcsms: BBC processes for verifying social media content

Below is the text of an article I wrote for the BBC College of Journalism website on verification processes for social media in BBC News, published 11 May 2011:

From Tunisia, via Egypt, to Libya and Syria, verifying and acquiring eyewitness/citizen journalist/user-generated content has become increasingly complicated as the material has become more sophisticated.

At the UGC (user-generated content) Hub in the BBC Newsroom in London, our process has become much more forensic in nature and includes:

– Referencing locations against maps and existing images from, in particular, geo-located ones.

– Working with our colleagues in BBC Arabic and BBC Monitoring to ascertain that accents and language are correct for the location.

– Searching for the original source of the upload/sequences as an indicator of date.

– Examining weather reports and shadows to confirm that the conditions shown fit with the claimed date and time.

– Maintaining lists of previously verified material to act as reference for colleagues covering the stories.

– Checking weaponry, vehicles and licence plates against those known for the given country.

That’s not an exhaustive list, but those are some of the most common things we do in BBC News.

This can take anything from seconds – as in the case of the eyewitness in Bahrain who pointed their laptop’s camera out of the window to prove they were near Pearl Roundabout – to hours, as we hunt for clues and confirmation.

During the battle for control of Zawiyah in March, taking our time, checking against existing sources and using our knowledge meant that we avoided running old material.

One clip appeared, linked to claims that it showed protests after Friday prayers on 4 March 2011. The location was one we had become familiar with. The clip appeared very soon after prayers and our suspicions were raised by the length of the shadows in some shots. The protester stripped to the waist and on their knees also seemed familiar to me.

Other outlets were running with it and there was a degree of pressure to verify it quickly. We soon had the answer when we found that one section had appeared in Ian Pannell’s report a week earlier.

We also made sure we alerted Libyafeb17.com, one of the main sources cited, to our findings so that it could reflect the correct information.

The expertise and local knowledge of both BBC Monitoring and BBC Arabic has been vital. They speak Arabic for one, and they understand accents, humour and nuance in a way we could never hope to match.

I can’t tell a western Libyan accent from an eastern one, but I know that there’s at least one colleague an email away who can. Likewise, when it comes to telling Homsi from Damascene dialect, there are people who can bring that knowledge to the process.

Around 3pm on 23 February, we were asked to verify video claiming to show bodies in the morgue of Al-Jalaa hospital in Benghazi. BBC Monitoring was able to confirm that the accents were eastern Libyan, what the captions and voices were saying, and to give context to the likely date of the events. By 6pm, this verified material was in the system in time for the Six O’Clock News on BBC1 and the BBC News Channel, as well as BBC World News and our website.

There are few organisations which can call on the breadth of resource and knowledge that we can within the BBC.

Whereas a single shot from an individual’s ‘perspective’ used to be the most common form we received, now more material appears as an edited sequence of clips, whether from traditional news agencies, activist groups or individuals.

Some activist groups, in the absence of foreign journalists on the ground, have filled the space left by the traditional agencies. Shaam News Network in Syria and Freedom Group in Misrata, Libya, are examples of how a motivated group with an internet connection can take ownership of the narrative. It’s important to remember the context in which they exist, especially in terms of awareness of their partisan nature.

But, regardless of provenance – from the most established source to the individual who has only ever uploaded one clip – everything we see goes through the verification process before we give our opinion on it. If it says it comes from the social media sphere, we question it until we’re happy that the claims being made stack up.

Contacting witnesses has also had to move with the times and circumstances. Monitoring and blocking of mobile phones and landlines mean that we’ve become increasingly dependent on VoIP (Voice over Internet Protocol) services, satellite internet and satellite phones to speak to eyewitnesses.

Establishing relationships of trust with people in such precarious situations is a huge challenge and one we take very seriously. We have very clear guidelines for contacting people, with the goal that we should never expose them unduly to danger. We will never ask or encourage anyone to put themselves at risk to tell us their story.

We are hugely indebted to these people, as well as family and friends who put us in contact with them. Without them, much of the story of the ‘Arab Spring’ would never have been told.

Categories
Digital Newsgathering Journalism Social Media Verification

Newsgathering in the open

Emily Bell has written an interesting piece on the notion of “live” as a defining characteristic of modern journalism. She argues that any news organisation worth its presence in the digital space needs to work in real time all the time and consider “live” as the default mode of operation.

In it she discusses the nature of Andy Carvin‘s ongoing experiment with journalism on the Twitter platform. I’ve discussed with Andy a bit about what he’s doing on twitter. This is the discussion we had about it back in March, in response to an article by Jay Rosen.