You can add value to your graphs by making use of Social Network Analysis – a group of methods and measures that highlight social relationships in network. This blog post gives an introduction to three of the centrality measures and the value they offer.

Please feel free to leave any feedback in the comments section below.

When to use Social Network Analytics

Before we can understand the different measures, it’s important to establish the types of networks where Social Network Analysis can bring value.

Given they are “social” measures, the networks they’re usually applied to are that involving the interaction of people. This could be a network displaying communications, transactions or even social relationships. The data used in this post is records of email communication at Enron.

Centrality, in the field of Network Analysis, is a concept that is used to identify the relative importance of a node in a given network. Here are the three types of centrality that KeyLines is able to measure:

Degree

Perhaps the most basic measure we use is Degree Centrality. This measure calculates which nodes have the most links in a given network. The more links a node has, the higher the value of degree centrality. This measure is vital in recognising important nodes, as it quickly highlights the players that broadcast and receive a high volume of information.

KeyLines comes fully equipped with a function that measures degree centrality for you. Nodes are resized according to their score, with larger nodes representing higher levels of degree centrality. This quickly makes important nodes visually identifiable in the network.

Betweenness

This is a centrality measure which identifies nodes that often appear on the shortest pathway of communication between two other nodes. In other words, high betweenness scores may relate to nodes that relay high volumes of information. In social networks, nodes with high betweenness are often located on the outskirts of two densely populated groups.

Someone with a high betweenness may be useful from a surveillance perspective, for example. If you were to allow your users to drill down to examine the individual data elements, someone with a high betweenness score would be a good place to start because they are likely to have information from disparate parts of the network.

The measure works in the same way, enlarging nodes which have the highest betweenness value.

Closeness

Closeness is a measure that indicates a node’s distance to all other nodes – therefore calculating its reach within the network. Someone with a high closeness score means that they are well connected throughout the network, and could be considered from a targeting perspective. Removing them from the network would cause significant communication problems.

In this diagram, KeyLines’ Closeness tool is demonstrated – larger nodes represent a higher closeness value.

KeyLines also has in-built functionality for measuring kCore, distance and shortest path – keep an eye on our blog in the coming weeks for further explanations of these analytic tools. If you’ve had experience of using SNA features or think that we’re missing anything, let us know in the comments section.

If you would like to know more about KeyLines, please contact us.

The post Social Network Analysis appeared first on .

You won’t hear about it on the news, but there’s an arms race taking place. Each year, an estimated $2.5 trillion is lost through fraud worldwide. Investigators and fraudsters are locked in a secret arms race to develop increasingly sophisticated tools and methods to outwit the other. With every new and costly technique that comes around – chip and pin payments, device fingerprinting, even telephone lie detectors – fraudsters inevitably find a way to get past them. But we want to show you how network visualization could be a huge leap forward in the fight against fraud, combining the power of computers with the intelligence of humans to add a whole new dimension to fraud detection.

The fraud balancing act

Existing fraud detection systems all basically work in the same way: huge volumes of data are collated, rule-scored and sorted into three piles: fraud, not fraud and unsure. In one insurance company we spoke to, those ‘unsures’ represented 2% of all claims made. It doesn’t sound like a big number, but it’s enough to make the manual review the most costly and time-consuming part of fraud detection. For some, that 2% may also represent the difference between making a profit and suffering a loss. The manual review is clearly a weak point for many companies fighting fraud; a careful balancing act between keeping genuine customers happy with fast, accurate decisions and preventing real frauds from getting through.

The fraudsters’ advantage

The problem is that fraudsters understand how these systems work, and they know how to exploit their inherent weaknesses. Organised criminals are experts at creating complex audit trails that data mining alone cannot detect. This, in turn, forces investigators to collect bigger and bigger mountains of data from every possible source – some accurate, some false, some intentionally misleading. With a broader dataset come more opportunities to find frauds amongst the false alarms, but without an efficient way to manage this data, an investigator faces an impossible task.

Visualizing Fraud: Uncover the relationships, uncover the fraud

Whether it’s a case of identity theft, account takeover or a false insurance claim, virtually all instances of fraud involve the fabrication of a relationship between people or entities. By modeling data in a way that highlights these relationships, investigators are able to highlight suspicious and fraudulent behavior.

Uncovering fraud with visualization

For an example, let’s consider an auto insurance claim… Mr B makes a claim on his insurance policy, saying someone drove into his car (Vehicle 1) at a traffic light. Your automated review process flags this as suspicious as Vehicle 1 was in an accident a few years back with Mr A, who also made a claim. An insurance fraud investigator looks at the claim and decides it’s legitimate. They’re two separate accidents in two separate locations involving two separate people. It’s just an unlucky car. What he misses completely is that Vehicle 1 is not only linked to Mr A and Mr B, but ALSO Mr C. Mr C and Mr B both live at 232 Main Street. Mr A and Mr B both make monthly insurance payments from the same bank account. Chances are Mr A, B and C are either all the same person OR three people working in cahoots. Also implicated is another car, Vehicle 2, which is registered to Mr C, who has been using it to make other fraudulent claims.

Confused?

Let’s look at the whole thing again, this time with a visualization. Mr B makes a claim for Vehicle 1, which was also claimed by Mr A:

We expand Vehicle 1 to see that it’s also covered by an insurance policy belonging to Mr C:

We expand again to see Mr A and Mr B share a bank account, and Mr B and Mr C share an address:

We expand one last time to discover Vehicle 2, lurking in the distance:

With a simple (and this is simple – we can expand into hundreds of nodes) we have managed to uncover three fraudulent insurance policies plus a vehicle that is most likely also involved in fraud.

The power of network visualization

It’s been proven that humans are able to interpret shapes and layouts more effectively than text or data. Integrating software like KeyLines into a case management system allows investigators to exploit their innate abilities and build a process that’s more fit for the task in hand, KeyLines can add whole new dimension to fraud detection. Instead of just checking for anomalies in expected patterns, or finding correlations in transaction characteristics, KeyLines links disparate pieces of data that might be four, five or six hops away from each other. Previously hidden or assumed irrelevant connections become new lines of inquiry and detection becomes faster and more effective. If you would like to know more about integrating KeyLines with your fraud management systems, get in touch for a demonstration or further information.

The post Fraud Visualization appeared first on .

When you’re building a network visualization component, ensuring the layouts are clear, consistent and good looking has got to be near the top of your list of objectives.

If you can’t see the patterns, correlations and networks in your data, what’s the point in a visualization?

Over the last few weeks we released a series of videos explaining how our layouts work. Hopefully this blog post will help summarise that information, and help you see which layouts will work for you.

If you have any questions, just leave a comment at the end and we’ll do our best to answer them.

Standard Layout

Standard is probably our most popular layout, being a good ‘catch-all’ solution for most kinds of network and any size dataset.

It works by creating a graph with consistent links with minimal overlap between nodes and edges so everything is distributed evenly around the chart.

The consistency and visual simplicity of the standard layout means it’s a good starting point for finding underlying patterns and symmetries before going on to explore in more depth.

Structural Layout

The Structural layout works in a similar way to the standard layout, but instead of distributing nodes around the chart, it groups similar nodes together. This is done by finding nodes with the same connective structure and fanning them around the parent.

This layout is useful for seeing which parts of the network interact with the same nodes – often indicating which nodes have the same role or perform the same task.

Radial Layout

This layout is great for identifying the nodes which are dependent on the function of others. It rearranges nodes in concentric circles around the subject node, with each ‘generation’ of node forming a new orbit outwards from the parent.

By using this layout you can clearly see which nodes directly interact with the subject node and understand how dependencies work throughout the network.

Hierarchical Layout

The hierarchy layout gives both a top-down and bottom-up view on selected nodes. Child nodes are shown in horizontal layers below the parents.

This layout is, unsurprisingly, mainly used for displaying hierarchical layouts with a strict chains of command, but equally can be applied to visualizing processes or workflows, as well as impact analysis.

Incremental Layout

KeyLines’ incremental layout function works in a slightly different way. It allows you to expand charts by introducing additional data, without affecting the elements already present on your chart.

This allows users to explore charts at their own pace and easily focus in on specific sections of the chart. As new data is added into the chart, KeyLines arranges it logically and clearly around the network.

Layout videos

To watch videos of all five automatic layouts in action, visit our Layouts page.

The post Effective Automatic Layouts appeared first on .

• Do you struggle to find insight in your data?
• Is your graph database more of a mystery than an asset?

If the answer to either of those is ‘yes’, then join us on Tuesday 01 October for our first ever Google Hangout - it’s free!).

In just half an hour, we’ll show you how visualization technologies can help you extract insight from your data and add real value in your organization.

Led by visualization expert, Corey Lanum, the Hangout will then explain how you can use KeyLines to to quickly build powerful, professional and scalable visualization applications that can be deployed throughout the organisation.

Date: Tuesday 01 October 2013
Time: 5pm BST / 12pm EDT / 9am PDT
Duration: 30 mins + Q&A

The post Network Visualization Hangout appeared first on .

We recently released a new version of KeyLines to our developers.

The latest version, 1.9.17, includes several new features and minor updates, designed to make KeyLines more flexible and the development experience simpler.

What’s changed in KeyLines 1.9.17?

• New background and foreground layers can be used to highlight certain nodes and ‘ghost’ others. This allows you to focus on what you need to see, without losing sight of the bigger picture. Our quick video shows how this can enhance your charts:

• Support for HTML5 dragging – you can now enable the dragging of items around the chart, customizing events when items interact with one another, helping to create a more intuitive experience. This is particularly useful if you’re using KeyLines as a drawing tool to build networks from scratch.

Minor updates:

• Customize font colors for selected items – this is no longer restricted just to white!
• Cancellable ‘mousewheel’ event – disable mouse wheel scrolling in your applications
• and minor bug fixes.

New demos:

We’ve built a few new demos for you to explore. Find them under ‘Demos’ in the Evaluation or SDK site.

• Chart saving and loading (‘Save and Load’)
• Resizing KeyLines dynamically on a page (‘Resize’)
• Background items (‘Background’)

The post KeyLines Release 1.9.17 appeared first on .

As KeyLines continues to grow and our customer base expands around the world, we have been fortunate to receive invites to speak at a wide range of events.

Here’s a quick run down of the Meetups, events and conferences at which we will be exhibiting and/or presenting in the coming weeks – we’d love to see some of you there:

• San Francisco – 04 October at GraphConnect
• Atlanta – 14 Oct at VizSec 2013
• London – 23 October at the CIFAS User Forum.
• Amsterdam – 24 October with Neo4J and Xebia
• New York – 05-06 November at GraphConnect
• Cambridge, UK – 12 November at our own friendly JavaScript Meetup
• London – 18-19 November at GraphConnect
• Norwich, UK – 21 November with our friends at SyncNorwich

The post A busy few months of events and meetups appeared first on .

Last week we welcomed our sixth new colleague to the KeyLines team.

Phil Rodgers (@PhilRodgers) joins us in the role of Software Developer, bringing with him more than two decades of expert research and experience.

Since graduating from Cambridge University, Phil’s career has seen him engage with a variety of cutting-edge software development – from desktop publishing to Formula 1 racing.

In his most recent position, Phil was part of the team developing i2’s Analyst’s Notebook product (now part of IBM).

Once the dust has settled, Phil will play an integral role in our research and development activity – working to further develop KeyLines and build new functionality that will cement its position as the most advanced network visualization toolkit available.

Phil is looking forward to working with a fast-growing new tool: “I’ve been really impressed with the work Joe and the team have done with KeyLines so far. It’s great to be working on this cutting-edge technology at such an exciting time in its development.”

The post A new KeyLines developer appeared first on .

We’re excited to announce that KeyLines was named the Most Innovative Graph Application in Visualisation at an the Graphie Awards in San Francisco!

The GraphConnect Awards, also known as The Graphies, recognise innovation and achievement within the growing graph database community. They are run by Neo Technologies, who make Neo4j – the world’s most popular graph database.

It’s fantastic to be part of such an exciting and innovative community, and a great honor to be recognized with a Graphie.

A full run-down of the GraphConnect Award winners can be found on Neo Technology’s website.

The post KeyLines wins innovation award appeared first on .

We recently released a new version of KeyLines, with new platform support, simpler packaging, a new demo, some API enhancements, and several fixes.

Simpler packaging

The keylines-html5.js file has been removed – its contents are now included in keylines.js.

IMPORTANT: The new version of keylines.js MUST NOT be loaded alongside an old version of keylines-html5.js. Please be sure to remove all references to keylines-html5.js from your web pages when upgrading to this version of KeyLines.

New platforms

KeyLines is now supported on Internet Explorer 11 and on iOS 7.

New demo

There’s a new Find Path demo that shows how to use the shortestPaths API to find the lowest-cost route through a network. Different weights on the links can give different results – try using the demo to navigate from Toronto to Washington by time and by distance.

API enhancements

The shortestPaths API now returns an extra property, ‘onePath’, in its results. This includes both the nodes and links in one of the shortest paths found. For compatibility, the ‘one’ property giving just the nodes is also returned as before.

Documentation

The “Hello World” example on the Getting Started page has been improved to give a more complete description of how to get started with KeyLines.

Bug fixes

• Arrows on links to combined nodes: When combining several nodes together, links to the combined node could sometimes have their arrows pointing in the wrong direction. This is now fixed.
• labelPosition: The labelPosition API now returns correct results for labels with more than one line of text.
• iOS 7 curved link drawing: We’ve worked around a problem introduced in iOS 7 which meant that curved links were drawn incorrectly.
• Layout improvements: Layouts have been improved to cope better with networks that contain a very large number of pieces that aren’t connected at all to the rest of the network.
• Special characters as item ids: In the Flash version of KeyLines, using certain unusual characters in item ids could cause problems in the graph API. This is fixed.

Please feel free to contact us if you have any questions or concerns. For the most up to date KeyLines information, make sure you’re following us on Twitter.

The post KeyLines release 1.11 appeared first on .

Yesterday the KeyLines ‘world tour‘ made its latest stop at the first ever GraphConnect London.

As you’d expect from a GraphConnect event, we had a great day hearing about the different ways people have incorporated graphs into their data and their work. We met with some fantastic people and answered a lot of different questions about graph and data visualization with KeyLines.

If you weren’t able to attend, you can view the slides below. If you’d also like to see the KeyLines demos from the presentation, just get in touch and we’d happily run you through them.

Thank you again to all the staff at Neo Technologies and GraphConnect for organising such a successful day!

The post KeyLines at GraphConnect London 2013 appeared first on .

Welcome to the first ever edition of KeyLines News – your run down of everything we’ve been doing here at KeyLines HQ.

The last 2 months

We’ve had a busy couple of months speaking with people all over the world interested in visualizing their data with KeyLines.

Buoyed by the success of our first Google Hangout (watch it here and keep an eye out for more soon) we attended a string of events from San Francisco to Amsterdam. Here just a few highlights:

• New York: Watch a video of Corey Lanum speak on interactive graph visualization
• London: See the slides from Joe’s presentation on terrible graph design
• Amsterdam: Learn about visualizing fraud with KeyLines, from our joint event with Neo Technologies
• Chicago: Corey speaking on the same subject in Chicago.

New faces

We’ve welcomed two new people to our team!

• Nate Benson joined as Sales Director and will be working closely with our customers in the EMEA region
• Phil Rodgers joined us from i2 and is already hard at work leading R&D on several exciting (but secret) new features.

A new website

Our public website, keylines.com, has been relaunched with a new design and new content. We’re adding new pages (and new downloads) all the time, so make sure you bookmark it and check back later.

KeyLines Version 1.11

Last month our customers received version 1.11 of KeyLines – now even simpler to implement, more compatible and with new features.

Get in touch if you’d like us to show you KeyLines in action, or to start a free evaluation of the toolkit.

Awards

Oh, and we also won a Graphie and were named a Global Game changer.

Not a bad couple of months. See you in 2014 for more!

The post KeyLines News – December 2013 appeared first on .

Being massive fans of the power of JavaScript we are proud to sponsor the NodeJS and JavaScript Cambridge meet-up group.

The group has 100+ members & we meet once a month in the upstairs bar of the Fountain pub to hear some of the luminaries and innovators of JavaScript speak.

Last night we were lucky to be joined by Mozilla’s James Burke – author of the well-known RequireJS module system. We are big fans of James’ proactive approach to getting modules working in JavaScript.  You are probably well aware that modules have not been part of the JS standards until very recently, with the push to get modules into ES6.

You can see the slides from his talk here: http://jrburke.com/talks/cambridge2013/  Thanks James!

If you’re in Cambridge in February, our next talk will be Jag Reehal speak on the subject of AngularJS – do join us

The post RequireJS at Cambridge meetup appeared first on .

On Monday we released KeyLines version 1.13.1. In addition to several minor bug fixes and incremental improvements, we released one cool new feature: Bubbles.

Introducing Bubbles

This new feature was developed following customer requests for a way to label nodes. They are designed to supplement glyphs and tooltips, which can still be used, by offering a higher profile text area.

Bubbles can be useful for so many different reasons:

• Adding unstructured data to your chart without ‘hidden’ tooltips.
• Adding more useful warnings to your nodes when a glyph (and its four character text limit) isn’t enough.
• Allowing users to supplement charts with annotations and notes
• Pretending Barack Obama endorses your product.

Bubbles can hold multi-line text and have no fixed character limit. The following attributes can be defined:

• Position (NE, SE, SW or NW)
• Bubble background color
• Bubble border color
• Bubble font size and color.

More information, including sample code, can be found in the SDK site under Documentation > Object Formats > Bubbles. You can also see bubbles in action using the ‘Styles’ demo.

For more information, get in touch.

The post KeyLines version 1.13.1 appeared first on .

Data visualizations aren’t just pretty pictures – they have a use. A well-designed visualization will help us understand data more quickly and effectively, which in turn leads to faster, more informed decision-making.

So, it can be frustrating when we see badly designed graphs.

Really badly designed graphs.

So rather than start ranting on Twitter, we thought we’d save our followers the ordeal and compile a list of basic rules to follow when creating your own graph visualizations.

If you have any you think we’ve missed, let us know in the comments section below.

Why visualize information as a graph?

Before starting, it’s useful to remind ourselves why some design elements work and others don’t.

Humans are not computers. Generally we find data intangible and difficult to process unless it’s encoded in a way our perceptive systems can quickly understand. Vision is our most powerful sense, with the highest bandwidth to the brain, so invariably the best way to encode data is visually.

We also want to exploit the brain’s ability to perceive ‘pre-attentively’, i.e. without conscious thought. For example, we can instantly understand that something smaller is ‘less’ than something large.

There’s been a lot of research into the visual cues most effective for this. You’ve probably already seen this kind of table in the information visualization literature:

To create an effective graph, we need to take these ideas, and some general design theory, and apply them to our work.

So here’s our…

10 rules of graph design

1. Start with the question – without knowing what you’re looking for, no amount of data or flashy graphics is going to help you. Start with a question and build your design around that. Ideally, stick to one question at a time, too.

2. Think about your colors – this applies of any kind of design. Some colors work well together; others do not. Think carefully about your color schemes, make your colors harmonious but also meaningful, avoid red/green combinations (not good for the colorblind) and generally any kind of rainbow palette is just going to disorientate.

   On a similar note, use a plain (and preferably white) background. Using an image behind the graph is only going to distract from your data, and black backgrounds should only be used if your audience is in a dark environment.

3. Avoid 3D – the unfortunate truth is that 3D graphs rarely work. They’re difficult to navigate without feeling seasick, suffer overlap problems and are almost impossible to print – even if you’re lucky enough to have access to a 3D printer.
4. Plan your labels – you have plenty of ways to add additional information to your graphs: tooltips, labels, menus, glyphs, annotation bubbles, color-coding, icons, animation, etc. Just remember that less is more; avoid overloading users by instead providing detail on demand.
5. Think about emphasis – depending on your question, you will probably want to emphasize certain nodes. Again, you have plenty of options for this – size, color, border, animation, glyphs, etc. Use them sparingly.
6. Easy navigation – allow your users to navigate the graph smoothly and intuitively. If you expect users will be using tablets or smartphones, full touch and gesture support should also be included.
7. Intuitive interaction – allowing your user to interact with the data helps them understand and discover new aspects of it. Each graph visualization should have a unique UI and UX, specific to the individual question it seeks to answer.
8. Use a good layout – a good layout will:
   • Leave minimal white space without bunching elements
   • Minimize node overlap
   • Minimize link crossings
   • Show symmetry and clusters where they actually exist
9. Make use of filters – Graphs are often large – in some cases too large for your users to see patterns clearly. Users need to decide which types of nodes and links they want to focus on, filtering out those that are less important.
10. Animate sparingly – it’s tempting to animate everything in sight. Resist the urge as it will overload users and limit the usefulness of your visualization.
11. Bonus tip: Use a good tool. We know one! Get in touch to find out more.

P.s. This blog post is based on Joe Parry’s talk at GraphConnect London – see the slides here.

The post The 10 rules of great graph design appeared first on .

Yesterday we made KeyLines 1.15.1 available to all of our current and evaluating customers. This new release has one new feature, a new demo and a huge number of minor improvements.

New feature: Retina Support

We’ve upgraded the way KeyLines renders graphics in Canvas, making them up to twice as sharp as previously. This means users of Apple Retina displays, or other high pixel density devices, will see KeyLines charts with crystal clarity.

It’s hard to show the difference this makes in a blog post – we’d recommend taking a look for yourself – but these screenshots give some idea of the effect (images zoomed to 150%):

New Demo: Titan Graph Database and Rexster

KeyLines has always offered easy connectivity with the Titan graph database, but our new demo will help users get started.

Using the demo that explores international trade data, developers can understand how KeyLines, Titan and the Rexster API fit together. Full source code can be explored and JSON objects are easily available for inspection.

If you have access to the KeyLines SDK, just navigate to demos > Titan to find the new demo.

To find out more about KeyLines and Titan, please visit our Titan page or download our guide; Visualizing the Titan Graph Database with KeyLines.

Minor changes

Since our last release in December, we’ve been busy making over 50 improvements, tweaks and bug fixes, including:

• Enhanced touch and gesture support
• Improved arrow drawing and labeling
• Smoother zooming and chart sizing
• Clearer (and more) documentation
• Bug fixes affecting users of IE11

If you have any questions about any of these changes, please do not hesitate to get in touch through support@keylines.com.

The post KeyLines version 1.15.1 appeared first on .

Do you know any budding data visualization developers?

This summer we’re offering the chance for a student (or recent graduate) to join the KeyLines development team.

Our Software Development Intern will gain ‘hands on’ experience of the software development process, including:

• Developing proof of concepts
• Building product demonstrators
• Carrying out Data investigation
• Integrating KeyLines with other products
• Product testing
• Maintaining the development environment.

Throughout the summer, the intern will be mentored by our team – the perfect kick-start to any data visualization career.

Is the internship paid?

Yes – we pay £1200 per month

Where is the internship?

The internship is based in Cambridge, so applicants should be eligible to work in the UK.

What would the working hours be?

We work about 40 hours per week, Monday to Friday. The intern will start in June for a minimum of 12 weeks over the summer holiday.

Is experience required?

Applicants do not need specific experience of working in a commercial software environment, but strong programming skills and an understanding of web technologies (especially HTML5 and JavaScript) are a must.

Additionally, to get the most out of the internship, applicants should be at least 2 years into a technical, scientific or mathematical degree at university.

How can I apply?

Send us an email (jobs@cambridge-intelligence.com) with your CV and a cover letter saying why you think you’d be a good fit.

If we think you sound right for the job, we’ll invite you to come and meet us and make sure you want to spend the summer with us. These meetings will take place in the spring.

Unfortunately, due to the volume of applications, we can only reply to successful candidates.

Are other internships available?

Yes! We also have a paid Market Development internship for summer 2014 – ideal for business, marketing or economics students. Find out more on our company website: http://cambridge-intelligence.com/jobs.

The post Software development internship appeared first on .

Welcome to KeyLines News – your digest of all things KeyLines. It’s been an exciting start to 2014!

Webinar: Visualizing the Neo4j Graph Database

Make sure you register to join our webinar with Neo4j VP, Peter Neubauer, and creator of KeyLines, Joe Parry.

The aim is to show you how to get started with your own full stack application for your graphs. Learn more and register here.

New releases

Since you last heard from us, KeyLines has jumped forwards two versions:

• KeyLines v1.13 – introducing Annotation bubbles, examples here.
• KeyLines v1.15 – pixel perfect Retina display support, comparison here.

Free resources: the Titan Graph Database

We’ve also improved our resources for KeyLines customers using Titan – including a new demo with sample source code, a getting started PDF guide and a video tying the two together. Visit our Titan page or get in touch if you have a Titan project you’d like to visualize with KeyLines.

Most popular content

Here’s a run down of our 5 most popular pieces of new content from 2014 so far:

• Ten rules of graph design – for graphs that work and look great.
• KeyLines for Social Networks – including a PDF download.
• Introduction to graph databases – what they are and when to use them.
• Case Studies – learn about current KeyLines deployments.
• Pinterest – See recordings of our talks plus videos and screenshots.

Events and meetings

We’ll be speaking at a few events in the coming months. Come and meet us if you can:

• II-SDV, Nice, France – 14 April.
• Neo4j Meet-up, London – 26 March, details TBC, Follow us for more.
• JavaScript Best Practices, Cambridge, UK – 28 February.
• NodeJS & JavaScript Cambridge, Cambridge, UK – 11 & 27 February, 11 March and 08 April.

Summer Interns

And finally, we’re offering two paid internships this summer – one in Software Development, the other in Market Development, perfect for someone looking to kickstart their career See job descriptions.

As always, feel free to contact us with any questions.

Best wishes

The KeyLines team

The post KeyLines News – February 2014 appeared first on .

Recently, more and more developers have been asking us if they can use KeyLines to visualize their Titan graph databases…

The answer, of course, is yes!

The KeyLines toolkit is flexible enough to work with virtually any datastore, including all graph databases. But it’s an especially good fit with Titan.

Over the past year, Titan has emerged as a great option for developers wanting a distributed open source graph database. Using Titan, you can store and query graphs of up to 100bn nodes.

To help Titan developers understand how KeyLines works, and what’s involved in getting a your own visualization application up and running, we’ve put together a collection of resources:

Getting started guide (PDF) – this guide includes an introduction to KeyLines, details of the architecture and sample code.

Download Guide

Visualizing Titan with KeyLines (video) – using KeyLines to visualize a Titan graph database – all summarized in 4m52.

Titan demo – if you already have access to the KeyLines SDK, you can access a demo application by navigating to Demos > Titan. Full source code is available for inspection.

If you would like more information on KeyLines, or would like to request an evaluation license to visualize the KeyLines SDK, get in touch!

Contact Us

The post Visualizing Titan Graph Databases appeared first on .

Earlier this week our North America Manager, Corey Lanum, co-presented a webinar with Neo4j VP of Community, Peter Neubauer.

The event was a great success with interest from attendees all over the world.

In case you missed it, here’s a recording of the webinar as well as a copy of Corey’s slide deck. We hope you find it useful – feel free to ask any questions in the comments section below!

The post Visualizing the Neo4j graph database with KeyLines appeared first on .

Today is our last day at ideaSpace – the Cambridge University enterprise accelerator we’ve called home since we founded three years ago.

The time has gone incredibly quickly, with thousands of companies, agencies and individuals contacting us from all over the world for help with their data visualization.

In that time, we’ve signed up new customers across five continents, opened an office in the US, talked at dozens of events worldwide and gained recognition (and even a few awards) along the way.

The KeyLines team has grown from just one person to seven and we plan to grow even more in the next year, so it was time to find our own place.

Our new offices are right in the centre of Cambridge overlooking Parker’s Piece. Feel free to drop in and see us if you’re ever at a loose end in the city.

As of Monday, our address will be:

Cambridge Intelligence,
Ravenscroft House,
59 – 61 Regent Street,
Cambridge,
CB2 1AB.

Our office number will also change to +44 (0)1223 362000.

Our legal registered address (115c Milton Road) and US office details remain unchanged.

We’d like to take a quick moment to thank everyone at ideaSpace – staff and fellow co-workers – for helping Cambridge Intelligence get off to such a great start!

The post Cambridge Intelligence scales up and moves out… appeared first on .