Here is the first in a series of session notes for KM World 2010 and Enterprise Search Summit 2010. I attended the workshop, Designing Social Search in the Enterprise. It was led by Thomas Vander Wal, Principal - InfoCloud Solutions, Inc, someone I have been following on Twitter for some time. So I was very interested in both the topic and the presenter. Please excuse the typos as these notes were done realtime with a quick edit. Here is the session description.
“One of the biggest problem on today's websites and intranets is findability. Even as large companies have scaled up their investment in search and intranet development, one key ingredient is missing: the people. People are social, and people leverage social strategies to find information all the time-at work, with colleagues-to solve real problems. How do we add this social element into enterprise search systems? How do we know what to look for in search tools, services to add into the mix, and or design elements that will make an impact?
Our experienced designer frames these relevant models of social search and walks though how to think through the problems to get to a more optimal solution. This will also include the potential benefits as well as the pitfalls that come along with some popular solutions (as well as how to counteract those pitfalls). You will learn to think and design through the considerations to augment findability with social search within the enterprise. He will discuss the tools necessary to understand search design, the benefits of social search, and how to work though real world problem sets interactively in the session, so those in the session will come away with the ability to apply what is learned to their work environment.”
Thomas began by saying that he avoids fuzzy terms such as community, expert, collaboration, and trust, as they are hard to define. So I guess I will not talk about building trust within communities to enable collaboration J. I certainly agree that these terms can lead you down blind alleys. A Google search on any of the tem shows the broad range of interpretation they can generate. I especially agree about the term expert. I generally feel that anyone who calls themselves an expert, has indicated that they are not one.
Enterprise search is problematic. It was an issue before social tools and has the potential to only get worse now with the new silos brought by some social tools. People cannot find stuff that they believe exists. One reason is that search often lacks the feedback element. Thomas mentioned two researchers, Marti Hearst and Bynn Evans, that have worked in social search. Thomas said he used some of their framing in his session. He now went through several types of social search.
There are several types of social search – friend/colleague filtered social search. Yahoo MyWeb 2 was an example but it is no longer available. There is a problem with this approach since personal networks may be narrow.
Social ranking is another type of social search. Digg, StumbleUpon and delicious are examples. The problem is that segmentation can be difficult.
Multi-person and collaborative search is a third example. System that support shared screens, such as Google Wave, is an example. The problem is that tools are limited. Google Wave is another tool that is no longer supported.
A fourth example is collective search such as Google News and Twitter trends. It can be used for exploratory search but the crowd is anonymous and cannot be filtered. There are also issues around shared vocabulary. One issue that Thomas found is fear of embarrassment because you are exposing yourself by using it.
Another example is massive scale Q&A tools such as Quora. It can be useful for handling difficult search problems but the problem now is that tools are limited, especially inside the enterprise.
Thomas discussed ways to fill in the gaps in social search. He offered four types of InfoClouds. Personal InfoClouds addresse the intranet. He used this concept to cover attracting things to your screen so they are available as you work. Then Thomas covered the concept of a Global InfoCloud, a larger set, but still within the enterprise. Then the External InfoCloud goes outside the enterprise. The Local InfoCloud completes coverage of the gaps. Information is often structured differently in specific locations such as a packy in Mass and party store in Michigan as a place to buy beer. The Local InfoCoud can have many components such as colleagues, social software, and portals.
As you use the other InfoClouds, you place content in your personal InfoCloud. Looking back at my days within a large consulting firm I can see how these work together. My hard drive was my personal InfoCloud. I looked to the other three for stuff to put in it, but I largely relied on personally stored content. The trouble was it was hard to search, as well as organize.
Thomas now looked at several tools, beginning with SocialCast. It is an enterprise micro-blogging tool that I have covered a bit (see Socialcast Adds Sharepoint and Outlook Integration, New Features, and Enhanced Metrics). They have a new tool called Reach. It drops micro-blogging streams within work tools. I like this as aligning social tools with work processes is generally a critical success factor. You can also save conversations so enterprise search can find them, another good idea.
System One is another tool. It has a semantic engine that structures the information as you type. It can fill in the gaps and bring up related information. You can search as you work, It also finds relevant people based on your content. Your entire document becomes you entire search engine. As you work, it searches and brings things up for your use.
The System One site says: “System One Radar tracks global information sources according to the context of your most important projects. Based on broadly definable Issue Portfolios it analyzes 80.000 news sources as well as Social Media, fully automated and in real-time, showing trends, backgrounds, relations, opportunities and risks. All visualized in a way that allows insight at a glance.” They have several tools and this is describing a Web version.
They also have a collaboration version. Here is its description: “System One Collaboration recombines the essentials of social software and semantic search in a highly integrated team productivity suite. Things that are being worked on are automatically related to all relevant resources within your company and beyond. System One Collaboration is browser based, works on mobile devices as well as on paper and is available as hosted or on-premise solution.”
James Robertson commented that these tools seem to attempting to make search support pervasive and come to your work. Thomas said that the trend is bringing these components to your existing tools so you do not have to learn new systems and they fit within your existing work processes.
Knowledge Plaza started as a social search engine and they added features over time (see my early review: Enterprise 2.0 Knowledge Sharing Platform from Knowledge Plaza). Thomas said he helped simplify their interface. It is a Web-based platform for enterprise search, social bookmarking, knowledge management, information brokerage and expert identification. It was developed by the Belgium based firm, Whatever. Nice name. There are facets, and you can search through other’s eyes. You can easily package the results and share them but it does not easily integrate into enterprise search engines.
Thomas next covered reconfiguring search with social relevance. He began with Yahoo MyWeb2 which was a bookmarking tool like delicious. You could search on others bookmarks. Thomas found it very useful as people he knew where curating for him by their own curation activities. A down side was that he could not contextualize the people, especially when they discuss off topic topics. Tagging could help here if people have the discipline to use tags. That is not a given. Some tools have auto-tagging that can help with this issue. Yahoo no longer seems to support MyWeb2. One reason is that Yahoo bought delicious.
Connectbeam is another example that unfortunately is not still around. I liked them (see Integrating Social Networking and Social Bookmarking with Enterprise Applications: Connectbeam). They believed that business networking begins around the sharing of ideas and information so they tightly integrated social bookmarking into their social networking platform. They also introduced the Connectbeam web services Application Programming Interface (API) that enables you to add full functionality of Connectbeam social software into your existing IT applications. Perhaps they were a bit ahead of their time.
Thomas next covered search across walled gardens. Enterprises create different levels of openness. Search needs to take this into consideration and honor the layers of accessibility.
Future search and follow results was the next topic. PubSub was an example of this but it is no longer around. You could get alerts on topics of interest for advanced business intelligence. Perhaps it was another company ahead of its time.
Thomas next discussed the concept of social comfort. There are three components: social comfort with others, social comfort with tools (e.g. How far is the reach of my actions?) and social comfort with subject matter (where can I talk about certain topics?). These are issues to resolve as you use social tools.
There was a lot of useful information and context in the session. Thomas has a depth of experience here. It seems that this is still a field that is a 'work in progress." Many of the initial tools are not around any more. Others are still being refined. One experimental tool in this space that I like is IBM Social Lens (see IBM’s Social Lens Provides Smart Content Filter). I look forward to seeing what progress is conveyed in next year's session.