Twitter search – finding old stuff

I was helping someone out with finding old stuff on twitter from a particular user. When I say old, I mean beyond the 3,200 limit of a single user’s tweets (that’s how many it stops at when you do that endless scrolling down on a page). I was also interested in searching hashtags used at particular times.

So this is really about searching between particular dates. The advanced Twitter search page doesn’t show how you can search for dates but you can do so by putting the following in your search:

  • from:username since:yyyy-mm-dd until:yyyy-mm-dd

So the url search for @stephenfry’s old tweets would be: https://twitter.com/search?q=from%3Astephenfry%20since%3A2009-10-01%20until%3A2009-12-01&src=typd&f=realtime

Or if you want to find a hashtag, word or phrase between particular dates:

  • #digitalbritain since:yyyy-mm-dd until:yyyy-mm-dd
  • digitalbritain since:yyyy-mm-dd until:yyyy-mm-dd
  • “digital britain” since:yyyy-mm-dd until:yyyy-mm-dd

You can also simply adjust the url instead of using the search box. For example, the urls for the above three searches (for dates from 1st October 2009 and until 1st December 2009):

In the url you should notice that:

  • %23 is a hashtag
  • %20 is a space
  • %3A is a :
  • f=realtime is the same as clicking ‘all’ in the search (gives you everything rather than ‘top’ tweets)

Hope this helps (it serves a reminder for me if nothing else). Welcome additions/corrections in the comments

 

Hyperlocal Publishing in the UK – A 2013 Snapshot

UK hyperlocal map

(Again, as soon as I get my login back I’ll cross-post this to creativecitizens.co.uk but if I don’t write it now I never will).

Last year I did some stats about how many news stories are produced by hyperlocal websites. I used the Openly Local database as the listing source and then counted the stories pushed through the RSS feeds of the sites (there was also a degree of tortuous manual counting as well. The method is described in a research paper – PDF). The big headline from that research was there’s a hyperlocal story published once every two minutes (during the day).

In summer 2013 I revisited the database and applied the same method of counting. Here are the findings. Some of these (mainly the geographic spread stuff) will be published by Ofcom in a report on ‘Internet Citizens’ in the next few weeks.

List of main findings:

Sample

  • The research draws on a list of 632 hyperlocal websites listed on the Openly Local database as of 7 June 2013.
  • 496 of these sites were ‘active’ and operating in the UK. ‘Active’ was defined as a website having posted a news story at least once in the 5 months prior to the sample period or functioned as a forum-only or wiki-based website.
  • 133 are no longer active. This figure is a mix of websites that have closed or have not published in the 5 months prior to the sample period.
  • The research sampled the news stories published by ‘active’ websites from 18-28 June 2013 (11 days).

Publishing

  • During the sampling period 3482 stories were produced by 224 sites (46% of ‘active sites). The number of stories produced in the 2012 sample was 3819 items by 313 sites.
  • The average number of posts of those sites that published in the 2013 sample was 15.5 posts per site within the sample period (12.2 in 2012). The median number of posts of those sites that published was 6 (7 in 2012).
  • 260 sites (54%) produced no story during the sample period (133 in 2012).
  • 38 (8%) sites produced just one story (39 in 2012).
  • 106 (22%) sites produced 5 or less items (133 in 2012). These were responsible for 8% of the posts (9.3% in 2012).
  • 87% of news stories (58% in 2012) were produced by 20% of the sites.

Commentary: the lower number of published sites is largely down to a reduction in output from ‘Local People’ sites . These sites have undergone radical change in the last 12 months with the large network of journalists writing for them now disbanded. The lower figure from these sites might also be accounted for in a refinement of the methodology, which attempted to avoid capturing adverts that were published in the news stream.

Frequency

  • Overall, an average of 13 items per hour were produced by Hyperlocal websites (15 in 2012).
  • During weekday daytimes this average rose to 22 items per hour (24 in 2012)

Geographic distribution

  • Number of sites in UK nations:
    England: 445 (+45 on 2012)
    Wales: 26 (+11)
    Scotland: 20 (+7)
    Northern Ireland: 3 (no change)
  • Number of sites in English regions:
    London: 96
    South West: 81
    South East: 73
    West Midlands: 59
    Yorkshire and Humberside: 39
    North West: 38
    East Midlands: 16
    East of England: 36
    North East: 7

Birmingham has 26 active sites (-2 on 2012) – the most in any UK authority area.

(A PDF summary of this. This report builds on data from the 2012 report ‘Hyperlocal Publishing in the UK – A Snapshot‘)

 

 

 

Asset Mapping Hyperlocal

Asset Mapping - #TAL13

[As soon as I get my login back I’ll cross-post this to creativecitizens.co.uk.]

Best to have a quick glance at the set of pics of this asset-mapping session on Flickr to give you an idea of what I’m talking about below.

Members of the Creative Citizens research team (myself, Jerome Turner and Andy Williams) attended an ‘unconference’ (#TAL13) on Saturday 29 September 2013, staged by our project partners Talk About Local. Having attended many rather more formal academic conferences throughout the year, this made for a a refreshing and invigorating change. Attendees at the Middlesborough event (hosted at MIMA) were drawn largely from the hyperlocal practitioner community with a sprinkling of members of the locally-based arts community.

Like the other strands in the Creative Citizens project we plan to make use of an asset-mapping methodology which aims to support community groups to consider what digital, human, physical or relational ‘assts’ they have to support their endeavour. It’s effectively a tool for reflection. At #TAL13 I wanted to pose the question:

‘What assets does your hyperlocal have?’

Initially I thought I might ask if it were possible to asset-map an ‘ideal’ or ‘sustainable’ hyperlocal, but that might result in participants trying to second-guess what I was implying by those terms so I left it more open and asked them to imagine their own hyperlocal operation in the centre of the map.

The detail of the method is outlined in a post from last year by Catherine Greene. Emma Agusita discusses how she used it with a community media organisation in Bristol.

In practical terms, participants put whatever they think more important closer to the centre of the map. Different shaped objects represent different things (spaces, infrastructure, media, groups and businesses, people, other). We had about six people contribute to our hyperlocal ‘map’ (there was a bit of drifting in and out). Here’s my take on what we found:

You don’t need an office, a nice café is handy, but a pub is essential.

Access to broadband was seen as essential but it didn’t necessarily matter where that access was. Cafés and pubs are good for “wifi, events, conversations” and it’s handy to have a “pub landlord who likes to be local hub of odd happenings”.

People with passion and a degree of skill are vital.

From “Paul, the web guru” to “John, our eager photographer”, you need people who have time to keep your hyperlocal ticking over technically and content-wise.

Public sector connections provide content.

It’s fair to say that many public sector organisations, from the police to local councils, now treat hyperlocals the same way they treat mainstream media. Even where they don’t, the degree to which the public sector has taken to Twitter means that hyperlocals are never short of access to stories or quotes.

Revenue matters / doesn’t matter.

This was placed both near the centre and at the edge of the map by two different people. We had Simon Perry from On the Wight with us who talked about needing to get serious about income-generation but others were less concerned about making money from their operation (why this was the case wasn’t made clear). It was noted that the same person who places revenue at the edge also placed ‘passion’ near the centre.

Other people’s stuff on the Internet provides content.

By which I mean, Creative Commons licensed images and Youtube. Several people mentioned trawling such sites for local images and discussed the value they create for them.

We don’t need apps.

In fact there were quite a lot of the ‘media’ icons placed around the edge. This partially represented a rejection of the need to engage with existing media but also a notable lack of interest in bespoke apps.

Thanks to all those who took part in the session. Any thoughts on this, let me know.