In the December issue of The Information Advisor’s quarterly knowledge management supplement, on an article on Twine and social graph searching, I wrote up an excerpt of an interview with John Tropea, a librarian and blogger at LibraryClips who has spent a great deal of time thinking and writing about social networks and graphs. I could only fit a short portion of the interview in our print publication—here is the entire interview, conducted by email: (apologies for the lack of embedded links!)
1. First, can you tell me a but about your background and blog?
In the information industry I have mainly worked as an academic librarian for an Australian college of natural medicine, several non-profit psychotherapy libraries, and about a year ago I moved from our Engineering corporate library to the document management team. As a corporate librarian the focus was on conducting research/analysis and delivery, in a way I was like a consultant working for many projects within our organization. Besides running the library the other main focus was a weekly blog of the latest industry news. Even though my new role is in document management I still can’t help sending people links to stuff I know (or hope) they like, if only everyone could have this librarian attitude, but not everyone is up to speed with who likes what in the enterprise or has the time, the librarian knows this as it’s part of their job.Jack Vinson, a KM consultant and blogger (Knowledge Jolt with Jack), blogged about this very aspect, in that knowledge sharing is not only sharing what we know but also sharing what we think is relevant to our colleagues needs. This would certainly be enhanced if people were connected in an online network, by looking at their content page, we would know what they like, and we would know the latest on what they are up to (the more you socialize the more you know about a person, the beauty is once you have added a contact to you network you can be updated on their activity stream, an insight into their daily world).Knowledge sharing is all about attitude and culture, but new technology can set the stage by allowing publishing, connections or networking to take place in such an effortless way. There is one big distributed pool of information, where “we” are the resources, where we tap into each other for information. This is in contrast to searching around for the right database and then messing about in that database. Instead we can now, in no time at all, find the right person to point us in the right direction as we are connected, and we may get more than a document, that is, we may easily connect with the author and learn more in depth and peripheral aspects of our task.Anyway being part of the document management team, gets me in bed with the tools that the knowledge management team advocate, and that’s exactly where I want to be.In relation to my blog, I started that nearly 3 years ago as a way to do something with all the information I was absorbing from the amount of feeds I was reading. Blogging and bookmarking was a way to remember all this stuff I came across, and of course the powerful stuff is that you are part of a blogosphere and discussions takes place, you can ask questions, so I find a blog as my voice. Blogs are part of an invisible type of community, my blog has comments and trackbacks from repeat visitors and bloggers, and I do the same with them and others. Blogs are not confined to the eyes of people registered with the network, they are just out there in the blogosphere and are indexed by search engines. But if the formal network is big enough blog networks or rather micro-blog networks like Twitter are extremely useful to learn off people and to ask questions. These days if I want to find something, I couldn’t be bothered searching I just ask the Twitter network and I could have an answer is seconds which could take me hours in searching, now I’m sure managers would like to hear that.2. What is your definition of social graph searching.?I see a social network as who you are tied to or connected to, and perhaps who they are connected to, all participants have an individual centric map derived from the same pool of people…basically it’s your friends and your friends’ friends.I see the social graph including: interactions, implicit intersections, and attention behavior.Interactions, are people that are not in your immediate network, but you have connected with each other somehow via the network or via a search result and then interacted via, private message, commented on their profile or post, etc…your paths have crossed, from there you could explicitly add them to your network.But more than this, using consultant and blogger Stowe Boyd’s term, it’s the implicit intersections or groupings you are a part of that you may not even be aware of. Once you are part of a network, you are an item in a database, and this item is grouped with other items, these groupings are slices of data. Show me all people in the network who live near me, show me all people in the network who are also interested in “libraries”, show me all people in the network who also speak Spanish. So you are part of these invisible groups or sets, and these people are part of your social graph, no different than the many boxes you fit in the demographics of the city you live in. I have something in common with this person as they live in my street, and something common with another person because we both like cooking. So this social graph provides opportunities for new relationships and information, a bit like recommendations. Other recommendations could be, these people read the same feeds as you, they bookmark the same sites as you, etc…
As Stephanie Booth, blogger of Climb to the Stars says, network systems have to take advantage of this by making these type of recommendations easy to execute, feed it to people so they can make new connections. But I’d also like to see more advanced ways of setting parameters for social graphing.
The recommendation part ties in with the social graph on attention data (reading behavior), here are people that read similar posts to you, here are people that also comment on this other person’s profile often, etc…
3. What’s an example of a social graph?
We get all this juicy information centered around the individual by default as long as we participate, explicitly share things, publish content and rove around the network, we get all this bonus information for no extra effort at all, which in turn offers discovery, new connection, new discussion, and perhaps more opportunity for innovation, I guess you could call this network effects.
The fun part is social graphing, more than your social graph, is being able to create your own social graphs. There is this talk of the difference between your social network and social graph, but can you not create social graphs within your network, only difference here is not about discovering new people, but more so the degree of closeness of your relationship.
Who in my social network comments on my profile the most; you can easily plot this on a graph.
But the main notion of a social graph is more about discovery, e.g. Show me people in the greater network (excluding my personal network) that bookmark the same sites as me, this allows us to find people with common interests that we didn’t know existed, and from this social graph you may add them to your social network.
You could do social graphing on the whole network (excluding or including) your network, you could even do it on your contacts networks (excluding the general network). It’s the kind of thing where a friend of yours is a friend of mine, well now you could graph this according to an interest, like who of my friends’ friends reads the same feeds as me. The exciting thing about this is being able to search your social filter, i.e. searching your friend’s profiles, even extended to your friends friends profile, or searching within blogs of people that read the same feeds as me.
I haven’t come across social graphing tools, but I heard of a tool that exports Facebook data from your network into a spreadsheet, with this you could create graphs. But this is limited to set parameters, instead we want to be able to create the parameters, i.e., slice the data ourselves.
E.g., Within my greater social graph who uses the same photo tags as me. People are on your x-axis and numbers on your y-axis. People would be arranged with the most similar person closest to the y-axis and so on.
I guess a more complex social graph could be done as concentric circles overlaid with a plus symbol, intersecting it into 4 quadrants The closer the person to the centre the more similar they are to you. Where they sit on the circumference of a circle is dependent on the quadrants. The top quadrants could be the northern part of the world and bottom the southern, and the left ones could be ages 0 – 40 and the right quadrants age 41+.
So a person who uses almost the same set of photo tags as you and lives in the southern part of the world and is over 41 years old would be plotted on a circle near the centre and sit in the bottom right position on the circle.
This type of graphing is only new in “perspective”, as it’s a graph based around “me”.
4. How do you feel it could be used within the enterprise as a way to better facilitate information discovery and sharing than current methods/approaches?
As I mentioned at the start social networks are only a more efficient and propagating way of doing what we already do, share information…we are sharing information all the time, we are always learning off each other without realizing it. With the perfect information sharing tools, like social networks, we would tend to share more often, have more people to share and learn from, and of course this is all searchable information for everyone to see. Like Toby Redshaw from Motorola mentions in a podcast, with social networks you simply magnify or multiply the chance of bumping into someone in the coffee room that can help you with your task, as now you are no longer limited to a geographic location and everyone has a profile and can be connected to in a flash. You can browse people or search the content leading you to people, and read up on what they are up to, what their thoughts are, what things they like, what is their expertise. You could request to add them to your network or just communicate with them.
Besides information sharing, the type or quality of information is more tacit, as web 2.0 tools allow you to publish, discuss thoughts and ideas so easily, you just wouldn’t share this quality of personal or spontaneous information in a document and upload it to a database, where no-one will know it exists. Instead we now have tools that harness this type of content, and it’s this type of personal content that is valuable know-how. Not only that, this content is alive, it can be discussed, and spore new content, in fact people get their own soapbox to share their thoughts, so it’s one big information organism that manifests and morphs. If the mailroom person had their space to share thoughts or was able to leave a comment on a managers blog, they could have the next big idea, at least for the department, where they are locals, and perhaps know best.
In the end, now that we know about these tools, it becomes an opportunity cost to not use them, this is the same for all things, but it’s double so when the opportunity cost is passing on something that itself opens up so much opportunity.
As you can see, to source information is easier when you can tap into an open people network, which can point you to information and perhaps fill you in with more insight that is on the document you are after.
If I’m after a document or about to create a document but want to see if something similar has been done, I can go to the network. Browse the expert and interest tags, search the network blogs and bookmarks, or blast a message to my personal network (maybe they can point me to someone).
Currently we search in the document management system, most of the times with no luck, as our search skills are deficient, or there lacks correct metadata or the document is located in a non-sensible area.
So then we email a closed set of people (mostly our hierarchy, like managers and colleagues), or ask people that sit next to us, or the guru in our office who seems to know everything.
With social networks you can perhaps find a blog post that points to a document, or even talks about the progress of a document. Blogs are good that way as they are just as much authored pieces as announcements and pointers to places. Or perhaps check who the likely people would be by looking at the expert tags, otherwise ask some people in your network.
This approach is much more connected and open, you are not on your own anymore, you are leveraging on the social capital of your organization. Plus documents that are being talked about give you context around the document, and the fact that they are worthy being talked about, so you may be pointed a good quality document amongst similar ones. The other thing is that since you have a network or a range of experts to look up, you find the right person much quicker, instead of the office guru answering questions all day, as they have the burden of knowing where stuff is and who knows stuff.
The other part to this is discussions build upon and shape ideas, and social networks allow this to happen, firstly it’s more open than email, everyone can see the discussion, secondly you don’t need to belong to a topic place to discuss, you can just do it from your own space. In saying this I still find forums a great way to discuss.
Anne Truitt Zelenka (consultant and blogger) posts about this scenario, adding that being connected to a network is not just another way of sourcing information other than being restricted to your hierarchy, but there is a chance you can find that information much quicker.
Lee Bryant (consultant and blogger) brings this to a more high-level saying that a network enables us to cultivate more autonomous behaviors. Alluding to, our connection with others, and our own space brings us closer to like people and the tasks that we are passionate about, projects that are of high interest tend to start gravitating to us.
Andrew McAfee (Harvard professor and blogger) has some great insight into network benefits, especially in the way of strong ties and weak ties.
Andrew mentions that your networks are made up of strong ties and weak ties, the weak ties are those people whose contact circle doesn’t overlap as much it does with your strong ties. I guess your weak ties are also people you have worked with before or share interests, but you don’t know everything about each other as much and interact as often as strong tie connections. When you are after some information, it’s your weak ties that can often help as they know people you don’t know, whereas your strong tie people often don’t bring anything to the table that you aren’t already connected to. Both are equally important, but if we just had strong ties, we wouldn’t learn anything new, it would become stale, not as much insight to innovation and other opportunities.
Dave Snowden (consultant and blogger) sums this up saying social networks and blogs are perfect for knowledge flow, as we are can share our thoughts whether it’s a need or not, i.e. in anticipation of our needs.
It’s not for this purpose that we share thoughts, we just to do it because of a sharing culture…if someone can use your past publishing’s to tackle a new need, then that’s great…what was a spontaneous published stream of consciousness is later being used to solve a problem or help with a task.
5. Have you seen any examples of firms that are doing this in anyway? If so, can you talk a bit about this?
I’m not in a consultant role and my current employer is about to launch communities of practice and blogs, so I’ll have a bit of first hand insight into KM 2.0. Of course the second phase of this would be to bring everything together in profiles or dashboards connected by a social network.
Luis Suarez’s (IBM consultant and blogger) blog is full of KM 2.0 in the enterprise; this is a great source for me.
6. Have you seen any software or services that facilitate social graph searching truly effectively? What do these do? How about within the enterprise?
Not really, maybe I just haven’t come across them. Fuser, a central site for all your email and social network inboxes has a feature called Leaderboard, where it shows who from your friends (your social network) talks to you the most, so this is more about closeness of relationships.
Facebook has a few widgets like finding how many degrees away from some you are; again this is within your social network. They also have a few other widgets like Friend Wheel and TouchGraph, these show you which of your friends are also friends. Socialistics goes a step further where you can map people in your network based on explicit data such as location, birth year, etc…
So I’m yet to see a widget that will let me see this type of graphing outside my social network, maybe because Facebook is a private system, whereas an enterprise version would have profiles as public by default.
Other than this there are the older tools which do an organizational network analysis or social network analysis based on interviews/questionnaires or mine data such as email and documents.
Social network participation makes this all available, instead of an analysis you just social graph it by clicking a button. This is limited to your online interactions, whereas SNA would also entail those physical interactions. A limitation of SNA is that the knowledge flows may not be important just because they are frequent.
Some social matching or analysis tools of late are: IntroNetworks, iQuest, NetAge, Trampoline, illumino.
Some of these are mining your content to find your interests, and then delivering new content based on these interests, it will also find the right experts to answer a question.
7. Are you familiar with “Twine”? What do you think its potential is for this kind of KM application?
No, I have just briefly read about it at the Read/Write Web blog
8. What do you think the best and most effective role would be for librarians and information professionals in helping organizations with social graph searching?
Start off by using the plethora of web 2.0 tools out there to connect people, a good example is Ning, where you create a community that has a forum, everyone gets a profile where you can add friends, message, and blog…another is Clearspace. Other tools are presence blogging like Twitter, and get people bookmarking using del.icio.us, get them reading feeds, using Yedda as an expert locator Q&A, perhaps using a startpage like Netvibes as a dashboard. Or perhaps request your enterprise as a special Facebook network, and you get all this in one. Social graphing comes later, for now we are just coming into the growth stage of social tools in the enterprise, culture and adoption are the focus at the moment.
9. What is your definition of KM 2.0, and how would it relate to this particular type of application?
KM 2.0 to me is the knowledge sharing part of knowledge management, having the tools and mostly cultural attitude to publish thoughts, ask questions, discuss and connect with each other in a network. Often emergent behaviors and patterns arise as the tools are freeform, non-structured…most of the times all you have is a title content, author, and category, and the content can be about whatever you want. These tools enable people to easily share things they normally would keep to themselves, by sharing we have a network effect, more opportunity for innovation and a network other than hierarchy as a source to get things done.
In regards to your social graph, it’s all about discovery, now you have a way to see people in your entire enterprise that have things in common with you, and then perhaps add them to your network. Usually this is done by browsing around or searching content leading you to a profile, but now we can graph these intersections.
Interview Links (not in order)
My blog links