Non-fiction Video Project Class @ Harvard (IXV)
Curiosity
How do we guide the users throughout the storied navigation activity?
How are we gonna invite them to explore small little new things, and investigate further based on the curiosity that has been aroused?
Can we build up a particular rhythm of navigation and, according to which, guide the users to come to one after another stage of the viewing activity, where the cinematic experience can be accumulatively elevated?
"Curiosity is one of the main driving forces that push people forward" is, in my perspective, universal in all kinds of storytelling activities. In the storied navigation activity where it is the users who actively determine how stories should progress, the questions that the system prompts should therefore foster the audiences' curiosity. The questions can include new information to the users (so it feels fresh), lead to sequences that provoke tension (e.g., pressure, discomfort, mystery, etc.) (so the story will be put into more energy and enhance users' engagement if paired with later relieves), lead to sequences that provide surprises (which will induce lots of users' own questions), and so on.
So, specifically, there're two questions that we should consider:
1) Are the questions that the system should ask what people tend to be most curious about? If so (I assume this to be true), how do we assess the curiousness that a question comprises?
2) How, what, and when, do we introduce new pieces of information that produces effects of new direction of navigation and/or tension?
Conceptual Relations in the Theme/Plans/Goals/Lessons
When editing a film, what the editors need to keep in mind all the time is, "What's the point am I trying to make right here/using this material?" They use footages that were shot at different time under different contexts and put them together as a sequence to make a point, as long as the semantics contributed in these footages can collaboratively build up the point that the sequence is meant to convey.
Often times, ordering sequences is something similar. When ordering sequences to make up a story thread, it tends to make great sense if sequences that convey related ideas are placed consecutively. Consider, as an example, this following sequence that I have for the film that I'm working on: Hugh is driving and talking to me, encouraging me to go to the PhD program. "You gotta do what's right," he said this punch line at the end of the sequence, with both his eyes looking toward me. It's a strong point that he makes, and shall at the same time be a strong point which I, as a filmmaker, make about his personality. That is: he has a strong idea about what's right, about one should do what's right for his/her own life, he is highly motivated about encouraging students, and so on. What are the sequences that we want to put behind this sequence? We can use the "sponsor week" sequence where he talks to the sponsors about the dream of making 100$-like artificial limbs, if we want to respond to "what's right"; we can use the "group meeting" sequence where he guides the students by encouraging them to explore the unknown, if we want to respond to "his motivation of encouraging students", etc. Anyway, all I'm trying to address here is, sequences are put adjacently because of reasons. These reasons are very much about, if not purely about, conceptual relationship.
In other words, if we can find conceptually related points that are made in certain sequences (presumably using the plans/ goals/ themes/ lessons indices by Schank), it makes a strong criterion of how the questions may be generated.
So, the third question we're to ask is:
3) How do we help users to input annotation in thesis indices, such that sequences can be identified as conceptual related? How is the system gonna generate questions, such that users navigation can proceed along the line of these conceptual relations?
延伸閱讀
posted by Edward Shen 沈育德 @ 10:28 AM
0 comments
