Posts

What next for the Web and information services? Linked data and semantic search

By Nicola Franklin

Where is the web going?  That was the question the speakers at the 2nd November 2011 NetIKX seminar were aiming to answer.  This join event, run with the Information for Energy Group (IFEG) and hosted at The Energy Institute, addressed the issue of linked data and the semantic web.

Whereas Web 1.0 might be thought of as ‘brochure ware’, one-way communication, and Web 2.0 has come to mean interactive, two-way communication online, the future seems to be for information and knowledge management itself to move onto the web.

The two speakers at this session described how this phenomenon is coming about, from two perspectives – Richard Wallis of Talis from the point of view of a producer or publisher of linked data onto the web, and Dr Victoria Uren of Aston University from the perspective of a researcher searching the semantic web.

What is linked data?  Richard gave an excellent introduction to the topic, leading us through a logical path to understanding how information from different data sets can be shared, merged and used online.  When the web originated, it was about publishing text documents with links to other text documents, using html.  Linked data is about linking ‘things’ to other ‘things’, by giving them a label or identifier (a URI).  Things also have attributes, like a name, size, location, etc.

The example Richard used was a spacecraft. 

A spacecraft is a ‘thing’ and can be given a label, such as:

1969-059A.

To make sure this is a unique label, some more information might be added, for example:

spacecraft/1969-059A.

To make sure people know it is your spacecraft you might add some extra information:

nasa.dataincubator.org/spacecraft/1969-059A

To store (publish) some information about this object on the internet you just at http://

http://nasa.dataincubator.org/spacecraft/1969-059A

When people start thinking about things, their attributes and how they link up together, they tend to think visually:

 

To transfer this into machine readable ‘computer speak’ the ovals are replaced by brackets:

<…/spacecraft/1969-059A>                    =  a thing

name Apollo 11 CSM                                       = an attribute and the value for that attribute

This language is called Resource Description Framework, or RDF for short.

Once common objects, or ‘things’, which are being  talked about by different people, in different locations, are identified by the same RDF label, then attributes or data about those things can be merged from those different sources – the data can be linked.

This can be very powerful.  For example, location data drawn from the Ordnance Survey can be linked with local authority data or central government data or NHS data.  This could answer questions like “how much was spent by this organisation in that area on this service, when this party was in power?”.

An example of linked data in action can be found on the BBC nature website.  This links together video archives from the BBC, information from Wikipedia, and information from other species or habitat-specific websites from various other organisations, displaying them all on one page.

Linked data can be used within an organisation, to publish data behind a firewall using intranet tools, which links together information from different business units, or held in different (perhaps incompatible) IT systems.  It can also be used to publish data externally to the internet, where other people and organisations can link it to their own data – either by using the same identifiers for ’things’ in common, or by mapping between their identifier and another one used for the same ‘thing’.

Some common standards are emerging, where ontologies or naming schemas are being published and adopted to ensure that different organisations use the same identifying labels to refer to the same ‘things’.  One example can be found at Schema.org, which is the standard being jointly adopted by Google, Bing and Yahoo.

How about semantic search?   Victoria’s talk began from the opposite end of the spectrum – given that linked data exists on the web, how do you search for it?

Traditional online searching is based around keyword search, which uses methods such as counting words, page ranking using links, controlled form searching (eg; OPAC) or metadata.  These methods were developed for searching text.  To search structured data needs a different approach.

Victoria listed a range of query languages that have been developed but said that SPARQL, which was based upon SQL, was the most widely utilised.  As it isn’t reasonable to expect users to familiarise themselves with a query language like this in order to carry out a search, a more friendly user interface is needed.

Again a range of methods have been developed:

  • Keywords
  • Forms
  • Graph based
  • Question answering
  • Tabular browsing

Victoria described the pros and cons of each method:

Keyword searching is easy to use but is restricted to simple searches for ‘a thing’.  Forms are a familiar interface, and allow more complex searches than single keywords, but forms need to be predefined and are therefore restrictive.  Graph based searches give a visual representation of the data, but this is hard to do for anything more than one ontology (ie, data from one source).  Natural language question answering is easy for the user, and good for heterogeneous data sets, but requires some heavy duty computing power to avoid being very slow.  Tabular browsing, where you start with one keyword and are presented with a whole range of linked words to chose from to narrow the search, can be clumsy.

Victoria felt that semantic search is very good for corporate data management, where information is typically focused around one topic area and it is very useful to be able to bridge between different data silos.  She gave examples of Drupal7, Virtuoso and Talis as systems that can be used for this.

Following a coffee break Syndicate Groups were set up to discuss several questions:

  • What is the value to a business of using linked data and semantic search?
  • Who would use the stuff from our organisation?
  • What are our needs for corporate data management – what tools are needed?

I took part in one of the two tables discussing the first question.  We felt that linking silos of information could help more people to find the right information, more quickly, and also to discover information they previously didn’t know existed (and therefore wouldn’t search for).  This could lead to finding the people behind the information and strengthening relationships.  It could also increase efficiency, raise cross-fertilisation and improve innovation.

During the group feedback session and discussion that followed, the issue of the risks of open and linked data was brought up.  Could increased ease of access to some data, and the linking together of many pieces of data from different sources into one location, be misued?  One example given was of insurance companies, potentially refusing to insure someone for a life or healthcare policy who they’d discovered had an unhealthy lifestyle.  Another example could be terrorists making use of combined information from Ordnance Survey data + google maps + other data sets to plan atrocities.

Linked data tools and open data publishing seems to have many potential benefits and also some risks; as with any rapid change the regulation and safeguards against the risks will probably lag behind what is taking place in practice.

Developing our capability – a seminar with Chris Collison

By Nicola Franklin

Yesterday afternoon saw me crossing London to attend my second NetIKX seminar, which I was looking forward to after a very interesting first foray a few months ago on the topic of social media use in organisations.  On this occasion we were to learn about the use of River Diagrams in facilitating knowledge sharing, from ChrisCollison, originally from BP and now working independently via Knowledgeable Ltd.

Chris introduced the subject by saying that River Diagrams are a way to visualise the results from a maturity model or self-assessment tool, looking through the lens of knowledge-sharing.  According to the OGC (* please see note below – Ed) a maturity model

“provides a systematic framework for carrying out benchmarking and performance improvement.”

Chris described knowledge management as a learning marketplace, with supply (of people with knowledge) and demand (of people needing to learn).  For any marketplace to work effectively, however, it needs a shared currency (so there is benefit to both sides in making the trade) and a common language (so that what is being traded is clearly understood by both parties).

The process of creating River Diagrams, and analysing them using Stairs Diagram (more on these later!) facilitates the creation of this common language as well as highlighting clearly which parties have knowledge and which parties need that knowledge.

OK, so what are River Diagrams and how do you go about creating one?

Firstly you need a group of participants.  These could be representatives from different departments within an organisation, people from several business units within a company,  or different stakeholders concerned with the same social, political or business issue.  Then you need to decide upon which topic or area of performance you want to benchmark or measure.

Chris described a River Diagram exercise from his time at BP, where they wanted to compare their 99 business units to agree which operational areas they all had in common (eg health & safety, corrosion management, water handling, etc), and benchmark performance of all the business units in each area.

First of all the participants needed to agree the operational areas and  ‘what success looks like’ for each one.  This is part of creating that common language.  To make the model work, they needed to agree five levels of performance for each operational area, from ‘world class’ (5) down to ‘basic’ (1).

This led to the creation of a self-assessment tool, looking something like this:

River Diagram Table

To fill this out, the participants from each business unit have a dialogue about where their unit falls, for each of the areas, until they agree a score for each one.  This results in a chart something like this, for this one business unit (let’s call them Group A):

River Diagram 1

Adding in the scores for the other business units increases the ‘width of the river’ or number of blue shaded cells:

River Diagram 2

With the remaining cells coloured green, the river analogy suddenly becomes clearer!

The ‘river banks’ represent those areas where none of the business units had a score.  Each group also records which two of the topic areas they would like to improve, and by how much.  Adding in this information allows the facilitators and participants to see where groups have knowledge to share (since they scored highly in that area) and where there are groups are keen to learn (since they logged a desire to improve their score in that area).

River Diagram 3

The black line shows the scores for Group A added to the composite diagram, and the two red lines show the two areas they chose in which they’d like to improve.

If you take one of the topic areas from the River Diagram, you can analyse the situation across all the groups in more detail for that topic by using a Stairs Diagram.  Here’s an imaginary Stairs Diagram for the ‘Corrosion’ topic area:

River Diagram 4

This shows each Group plotted according to the levels they scored (vertically) and the gap between their current level and the score they would like to reach (horizontal).  You can see that Group A could benefit from having a dialogue with Group B or E who are both at level 5 for this topic.

Chris went on to tell us about using the River Diagram technique to help the UN HIV and Aids group work with stakeholder groups across the world, to find out and benchmark the elements that make up a successful Aids management programme and to help the different groups learn from each other.

Chris emphasised that a lot of the work for this technique goes into the selection of topics and discussion to agree on the descriptions of each of the different levels, for each topic.

Once he’d explained how it all worked, it was time for us to have a go at creating our own River Diagram.  Since there wasn’t time in the afternoon session to write all the level descriptions, Chris had pre-prepared one based on a topic he hoped we were all familiar with – information and knowledge management!

He asked us to rate our own organisations for each of the topic areas he’d selected (eg, Knowledge Strategy, Using and Accessing Expertise, Exploiting Information, etc). It became clear that there would be a lot of value in getting different stakeholders in an organisation, department or other group to discuss where they felt their body scored for each area – and why.

One question that came up was, if one group’s goal was to improve in a topic, but none of the other groups had scored highly in that area, how could you improve ‘beyond the river’ (ie into an area on the ‘north bank’ that is coloured green)?  Several suggestions were made:

  • Bring in an external consultant
  • Set up peer referencing with another organisation (in your industry or from a completely different field)
  • Generate your own innovative solutions internally

Another suggestion from the floor was that it would be valuable to re-run an exercise like this in 6 months time, to get a comparison and measure whether any improvements had been made in the targeted areas.

A final thought that Chris added to the group discussion were two questions knowledge managers should ask the leaders in their organisation to ask all the time:

Of people with a problem = “who can you learn from?”

Of people with a success = “who can you share this with?”

The formal part of the afternoon concluded at about 5.00pm, followed by some equally enjoyable wine and networking.  I found this a very enjoyable and interesting session, and think this would be a valuable tool to add to any information and knowledge manager’s armoury.

Nicola Franklin

Director, The Library Career Centre Ltd

P.S. Chris also suggested these 2 links might be of interest:

http://www.communitylifecompetence.org/en/29-aids-competence

http://www.youtube.com/watch?v=Ai7LcCzOJo8&feature=fvst

*Note from the Editor (with thanks to Graham Robertson)

The OGC website will cease to exist from 1st October 2011Any new information will be published on the Cabinet office website:
http://www.cabinetoffice.gov.uk/All information currently on the OGC website will be available on the National archives website:

http://webarchive.nationalarchives.gov.uk/20100503135839/http://www.ogc.gov.uk/index.asp

Please update any bookmarks you may have, and if you have any queries or questions please contact 
==================== ++++++++++ ====================

Knowledge management and organisational strategy – NetIKX March 2011 seminar

Another lively and thought-provoking NetIKX seminar took place on Thursday 24 March at the British Dental Association which is fast becoming our preferred location as healthy numbers of people continue to support our programme. Flooding at Trafalgar Square, power cuts on the underground and speakers stuck on trains gave us a few hairy moments but in the end nearly 40 participants enjoyed 2 different but very good speakers, Dr Nick Milton (not Wilton as per the feedback sheets!) from Knoco Ltd and Linda Wishart from the Department of Health, followed by syndicate work on Nick’s Boston boxes, Working across cultures, Exit interviews and What is KM all about?. The overriding message was that KM is necessary now more than ever.

Nick offered us a model of knowledge based on competence in 4 different areas – Potential competence (in business terms looking at emerging markets), Competitive competence, Core competence and Others’ competence. The “Others” box led to some interesting discussions during a brief Q & A session at the end of his presentation. Nick’s view was that this led to Outsourcing and Quality assurance but he admitted that his views were likely to change as a result of the debate! It is good to see NetIKX influencing the minds of today!

He then gave us 2 stories – the first on a successful deployment of KM principles leading from Strategy to Activity and Results. Although I never got to understand how Mars chocolate can stop melting in India, the successful outcome was that sales trebled and the profit % doubled. Nick then offered us a cautionary tale when these principles are not followed. It was a sobering scenario about the disaster at Longford refinery in Australia which was caused by a knowledge failure resulting in 2 deaths and 8 people injured as well as a loss of power in the area for 20 days!

Linda’s presentation on the challenge of implementing KM in changing times when central government is heading towards significantly reduced workforce and resources, shared services and no money, Information Assurance and risk management, coupled with the government’s transparency agenda, was all too familiar to colleagues from other government departments. Linda did admit that DH was the first department to be taken to task by the Information Commissioner who had concluded that their FOI requests and record management was not up to scratch. After providing some insight into the myriad of different roles which the DH Knowledge worker will be facing in 2015, Linda shared her strategy of facing up to the KM challenge. This strategy was based around improved technology where possible, improved information access, knowledge capture and transfer, training and awareness, and engagement with workgroups. Linda was unfortunately unable to stay for the syndicate work but she was able to answer a couple of questions. Very similar to Nick’s premise that KM has never been more important, in response to the question “How does FOI impact on people’s willingness to record knowledge?”, the answer was that it is all about managing information properly. So there you are then – Knowledge Management – does it still have a role in organisational strategy? The answer was a resounding yes!

There were lively discussions at the 4 syndicate groups and the report back was interesting and well-received.

Melanie Harris

Using social media to achieve organisational goals – the next steps

Blog by Elisabeth Goodman

A shift from skepticism about, to evangelism for Social Media?

On 19th January, NetIKX hosted what proved to be a very successful seminar on this theme, with speakers Dr Hazel Hall1 and Nicky Whitsed2.   It was a follow-on seminar to one hosted the previous year, where we had introduced our members to a range of social media tools, and questioned if and how NetIKX might use them and also guide people in their use3.

Although our January 2010 seminar was also very popular, there was still some skepticism about the value of social media tools, and how organisations might use them.  This time, as Hazel commented to me in an aside at the end of the meeting, the tone was perhaps more one of how organisations might be persuaded to adopt the wider use of social media.

Social Media can be used by Library and Information Departments for a diverse range of purposes

Our speakers described the wide range of uses that social media tools can be put to, and their ability, beyond that of the previous tools available to us, to connect people as well as data and information.  We and our customers, can use social media tools for:

  • Collaborating on projects and for learning through wikis and ‘tweet-ups’
  • For staff development, teaching and training e.g. through ‘amplified events’ where someone present at an event will be sharing the content through Twitter with those who cannot attend.  Or by posting a recording of the event for others to access afterwards.  The Open University use Illuminate to run and record such events.
  • Providing virtual reference sources
  • Seeking feedback or peer review on planned presentations (which Hazel did for this presentation)
  • For gaining a better understanding of customer needs leading to new service developments

As Nicky pointed out, it’s important to understand the tools that our customers are using, and to be able to deliver services through those.  In fact her department has a ‘digilab’ where they have all the latest technology and social media tools, enabling their staff to become familiar with their use, and experiment with new ways of delivering their services.

The adoption of Social Media will be evolutionary, with some people leading the way

In the syndicate discussion groups that followed the presentations, delegates discussed the already visible evolutionary pathway in the adoption of social media by organisations.

Human Resources departments are using tools such as LinkedIn to learn about potential recruits.

Sales and Marketing teams are using Twitter and monitoring the web to find out and in some cases respond to what their customers are saying, monitor the competition and also influence the perception of their organisation.

Some companies are using tools such as Yammer internally to try out the use of such tools, or even to support the ‘crowd-sourcing’ of ideas in project management or general problem resolution4.

There needs to be a fine balance between policies and trust

It’s certain that organisations need some form of policy for the use of social media to address such issues as security and ethical behaviour.  Nicky shared details of sites such as http:/socialmediagovernance.com that can help us with that.  However, policies need to allow sufficient scope so as not to discourage the use of social media.

Library and Information professionals could influence the policies within organisations, and even encourage the adoption of values or competencies within performance review frameworks that promote knowledge sharing through social media tools.

As we discussed in one of the syndicate groups, people are used to assessing and building trust through face-to-face interactions.  Social media users are now finding proxies for building that trust, for example by relying on the judgment of those whom they know already, seeing which postings are re-tweeted by others, reviewing the posting history of new people that they ‘meet’ online.

Increased adoption of social media by organisations will require a cultural change

Again, as put by one of the syndicate groups, we are operating in a ‘perpetual beta’ environment.  This is a shift for organisations that are used to making decisions on well-established software with a firm support infrastructure.

As Hazel put it, we also have a ‘youngster elders’ scenario, where people who are perhaps more used to leading and being the authority on subjects, need to be open to seeking guidance from the more knowledgeable younger generation (as some of us may already be doing at home!).

Hazel and Nicky described how Library and Information professionals can play a role in guiding and supporting the evolutionary adoption of social media tools by:

  • Demonstrating how the tools can be used
  • Experimenting and developing our own capabilities, as well as giving our users the opportunity to experiment
  • Providing training e.g. in digital literacy

Concluding thoughts

The use of social media tools in the organisation should be part of Library and Information Management strategy but they tend to be owned by Security.  We need to help organisations to switch from an emphasis on the risk of using social media, to the risk of not using these tools.

Notes

  1. Dr Hazel Hall is Director of the Centre for Social Informatics in the School of Computing at Edinburgh Napier University. She is also leads the implementation of the UK Library and Information Science Research Coalition. Hazel was named IWR Information Professional of the Year in December 2009.
  2. Nicky Whitsed is Director of Library Services at the Open University.  She is an experienced strategic and change manager having led successful projects in the commercial, medical and academic fields. Nicky is trained in project management and facilitation and also has experience as a trainer. She has served on a number of CILIP and JISC committees and on a number of editorial boards.
  3. Elisabeth Goodman and Suzanne Burge presented on ‘Social networking tools – should they be taken seriously’ in January 2010.  See Elisabeth’s presentation: “Using LinkedIn, blogs and Twitter for networking and communities of interest”
  4. See related blog by Matthew Loxton on crowd-sourcing
  5. Whilst writing this blog, several of the participants at the seminar also shared their accounts of the meeting.See for example the following:
  6. Elisabeth Goodman is the Programme Events Manager for NetIKX, and is also the Owner and Principal Consultant at RiverRhee Consulting, providing 1:1 guidance, training / workshops and support for enhancing team effectiveness through process improvement, knowledge and change management. She also provides 1:1 tutorials, seminars and workshops on the use of LinkedIn and other social media. Read Elisabeth Goodman’s blog for more discussions on topics covered by this blog.