Notice the intense look of quietly attentive search on the upturned faces of the boy and of the man in the photo above. What are they looking for? Do they know — exactly — what they are attempting to see, or to learn? Or are they — at least in part — discovering what it is that they are seeking through their looking itself?
Searching for information, or seeking for ideas, can often be like this. We may have a sense of the general direction in which we should be looking, yet not quite know exactly what it is we seek.
The early starts and steps of our creative projects, too, can be like this. We may have a vague feeling of the direction we'd like our project to explore. It may be that we know that a new image or text should be, for example, dynamic, colorful and energetically strong, or layered, evocative and subtle. But we may not yet know how to move from these general and vague goals to our next steps. We're seeking, we're actively searching, but are unsure how to proceed. We are exploring, often in unfamiliar or barely familiar terrain where we don't want to miss that rare or unexpected treasure.
What to do?
One promising computerized approach using an innovative form of reinforcement-based learning is called "interactive intent modeling." This new approach is grounded in the conceptual connections and meaning networks between topics and subtopics themselves.
What if your search query outcomes were displayed on a large radar-like screen where you could physically swipe with your finger the best results to the center of the screen? What if you could move less relevant options to the screen's edge, through a tactile, touch-based interaction that visualizes where you are in search space. And what if, once you've spatially grouped some results to the center, the computer refreshes itself, proposing new relevant options that you hadn't thought of, and makes new suggestions based on what it has learned about your intentions. And the cycle repeats, with you dynamically learning and discovering as you go.
A team of researchers from Helsinki, Finland, created just such a computer-based search interface for a database of more than 50 million scientific articles. In one study, participants were asked to imagine that they were to write a scientific paper on a designated topic, such as "robotics." All of the participants were students in Masters or PhD programs and so were generally familiar with scientific research processes but not with the specific topic they were asked to write about. The researchers then compared performance of one group of participants given a conventional list-based typed queries, with a second group of participants using the radar-screen-like interactive visualization of their search.
Across the session, the interactive intent interface resulted in a significantly larger number of relevant documents being retrieved. The retrieved documents were also higher in novelty, yet still included as many obvious (central or classic) papers as found through the conventional list-based search interface. These increases did not add to the overall search time. Instead, participants using the interactive intent interface had more frequent iterations or cycles of search in the same period of time than did those searching with a conventional line-by-line "typed query" method.
Asked about their experiences at the end of the experimental session, many participants pointed to how the interactive keyword visualization was easy to use, noting that it helped them to identify related keywords that they didn't know. It also provided new cues even for well-known topics, and allowed them to quickly and easily emphasize some directions of search while de-emphasizing others.
Some searching questions
Although computerized interactive search where the computer comes to "divine our intentions" is one promising way to improve the likelihood that our search for inspiration will yield what we hope, the lessons we learn from this can help us in other forms and modes of inspirational search. Here are some questions for you to think about.
How purposeful or active is your search? When embarking on a new project, or a new subpart of a project, do you:
Where and what are your "treasure troves" for inspiration? Could you organize, group, tag, or categorize them differently so they would better meet your individual ways of searching? Do you juxtapose, interpose, or re-shuffle your image/word treasure troves from different projects, or keep them separated from each other? Have you tried both mingling and keeping them separate? What works better for you, and why?
How could you adopt the central ideas of "interactive intent search" –– creating a search space that adaptively responds to not only your first directly stated intentions, but also your emergent indirect and dynamically changing intentions –– to better display and share ideas, with yourself at different times, or with others, to increase your "inspirational odds"?