Google recently unveiled its public data search that allows people to generate charts of data from public sources. Nifty. But it doesn’t seem to allow the user to customize the visualization by selecting representations, date ranges, etc., or for data to be extracted for further analysis. It would be great if I could run my own statistical analyses on the data, or generate visualizations with, for example, Many Eyes.

Another problem I see with the data is lack of transparency: I couldn’t see any way to browse the various datasets they have indexed. Instead, it seems that you have to stumble onto them by chance. Nice for serendipity, not so nice for exploratory search.

I saw an interesting bit of technology at CHI 2009 this year. Yevgeniy Medynskiy, Mira Dontcheva, and Steven Drucker published a paper called “Exploring Websites through Contextual Facets” where they tried to solve the problem of iterative query formulation in online faceted search. They observed that search interfaces that allowed the user to specify multiple criteria to find desired objects often removed the search interface when the user selected item details. In addition, they observed vocabulary mismatch for aspects in the search interface vs. the details view. Finally, people often landed on item pages through external search engines, and couldn’t easily get to an in-site search interface with their search context preserved.

Continue Reading

A paper presented at CHI 2009 described strategies and processes used by intelligence analysts. Among other aspects, the paper discusses collaboration among analysts, quoting one of their participants:

What I will not trust and put into my analysis is somebody else’s analysis. I need to know the source of the information  and build on that so that I can put my level of trust in it and then it’s my name at stake when I provide an answer… I won’t trust their analysis until I look at the source of the information, and it will be, “Do I agree with the conclusions that they came to based on the facts and the evidence?”

Continue Reading

The area of collaborative search has experienced significant growth over the past couple of years in the number of research groups interested in the topic, and in the number of research papers being published in proceedings of respected conferences.

Interestingly, there is also a rash of “best paper” awards for this work. In chronological order of publication, the following papers related to collaborative search have received “best paper” designations.

• Pickens, J., Golovchinsky, G., Shah, C., Qvarfordt, P., and Back, M. (2008) Algorithmic mediation for collaborative exploratory search. In Proc. ACM SIGIR ’08. ACM, New York, NY, 315-322. (Best paper award)
• Joho, H., Hannah, D., and Jose, J. M. (2008) Comparing collaborative and independent search in a recall-oriented task. In: Proc. IIiX 2008, 89-96. (Best paper award)
• Paul, S. and Morris, M.R. (2009) CoSense: Enhancing Sensemaking for Collaborative Web Search. Proceedings of CHI 2009. (Best Paper nominee)
  

Congratulations to all the authors, and please let me know if I forgot to list your paper!

In their recent ECIR paper, Joho et al. explored the effectiveness of several mediation techniques around relevance feedback in supporting collaborative search. They ran simulations based on TREC HARD topics queries elicited in an earlier study, and found that certain techniques were effective at increasing recall during (simulated) search sessions consisting of 20 queries.

Continue Reading

Giles Crouch wrote recently about social search vs. general search, pointing out that people often search for information in their social network vs. in a general index such a Google or Yahoo! While we need to distinguish the cases when people search their social network for information about the network per se vs. information that the network refers to, there are circumstances when people make explicit decisions about where to search. It seems to me that there are several reasons why people may prefer not to use generic search in certain cases. Continue Reading

Late last year Hideo Joho, David Hannah and Joemon M. Jose published a paper that described an experiment in collaborative exploratory search. They compared teams of pairs of searchers in three conditions — independent (not collaborative), collaborative without communication, and collaborative with communication. This paper is interesting for several reasons, not least of which is that it made an attempt to quantify the effects of collaboration on search performance, an important subject that has not yet received adequate attention.

Continue Reading

At CHI 2009 this year, Sharoda Paul presented a paper she co-wrote with Merrie Morris that explores how sensemaking can be managed in a collaborative search environment. They created CoSense, an interface that augments SearchTogether with several tools that facilitate awareness and information sharing among collaborators. Tools include interactive query timelines, statistics on individual queries and term use, chat history, and a workspace for annotating search results.

Continue Reading

There was a good crop of papers at CHI 2009 this year, and I didn’t get to see them all. I did see a few that were particularly interesting, including   “Learning How: The Search for Craft Knowledge on the Internet” by Torrey, Churchill, and McDonald. The paper describes and analyzes search activities by people involved in various crafts. This work is interesting to me because in a way it very clearly separates exploratory search from other kinds of online searching.

Continue Reading

Thus far in our series on Collaborative Information Seeking we have explored two dimensions: Intent and Synchronization. The next dimension is the Depth at which the mediation (aka support, facilitation) of the multi-user search process occurs.

We can talk about three levels of mediation: communications tools independent of the search engine (e.g., chat, e-mail, voice, etc.), UI-level mediation, and algorithmic mediation. The first level typifies most searching currently being performed on the web, whereas the other two are more commonly found in research prototypes. Continue Reading