An official website of the United States government
Interactive textbooks generate big data through student reading participation, including animations, question sets, and auto-graded homework. Animations are multi-step, dynamic visuals with text captions. By dividing new content into smaller chunks of information, student engagement is expected to be high, which aligns with tenets of cognitive load theory. Specifically, students’ clicks are recorded and measure usage, completion, and view time per step and for entire animations. Animation usage data from an interactive textbook for a chemical engineering course in Material and Energy Balances accounts for 60,000 animation views across 140+ unique animations. Data collected across five cohorts between 2016 and 2020 used various metrics to capture animation usage including watch and re-watch rates as well as the length of animation views. Variations in view rate and time were examined across content, parsed by book chapter, and five animation characterizations (Concept, Derivation, Figures and Plots, Physical World, and Spreadsheets). Important findings include: 1) Animation views were at or above 100% for all chapters and cohorts, 2) Median view time varies from 22 s (2-step) to 59 s (6-step) - a reasonable attention span for students and cognitive load, 3) Median view time for animations characterized as Derivation was the longest (40 s) compared to Physical World animations, which resulted in the shortest time (20 s).
more »
« less
Gemini 2 : Generating Keyframe-Oriented Animated Transitions Between Statistical Graphics
Complex animated transitions may be easier to understand when divided into separate, consecutive stages. However, effective staging requires careful attention to both animation semantics and timing parameters. We present Gemini^2, a system for creating staged animations from a sequence of chart keyframes. Given only a start state and an end state, Gemini^2 can automatically recommend intermediate keyframes for designers to consider. The Gemini^2 recommendation engine leverages Gemini, our prior work, and GraphScape to itemize the given complex change into semantic edit operations and to recombine operations into stages with a guided order for clearly conveying the semantics. To evaluate Gemini^2's recommendations, we conducted a human-subject study in which participants ranked recommended animations from both Gemini^2 and Gemini. We find that Gemini^2's animation recommendation ranking is well aligned with subjects' preferences, and Gemini^2 can recommend favorable animations that Gemini cannot support.
more »
« less
- Award ID(s):
- 1907399
- PAR ID:
- 10355037
- Date Published:
- Journal Name:
- IEEE Visualization Conference (VIS)
- Page Range / eLocation ID:
- 201 to 205
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
We present Animated Vega-Lite, a set of extensions to Vega-Lite that model animated visualizations as time-varying data queries. In contrast to alternate approaches for specifying animated visualizations, which prize a highly expressive design space, Animated Vega-Lite prioritizes unifying animation with the language's existing abstractions for static and interactive visualizations to enable authors to smoothly move between or combine these modalities. Thus, to compose animation with static visualizations, we represent time as an encoding channel. Time encodings map a data field to animation keyframes, providing a lightweight specification for animations without interaction. To compose animation and interaction, we also represent time as an event stream; Vega-Lite selections, which provide dynamic data queries, are now driven not only by input events but by timer ticks as well. We evaluate the expressiveness of our approach through a gallery of diverse examples that demonstrate coverage over taxonomies of both interaction and animation. We also critically reflect on the conceptual affordances and limitations of our contribution by interviewing five expert developers of existing animation grammars. These reflections highlight the key motivating role of in-the-wild examples, and identify three central tradeoffs: the language design process, the types of animated transitions supported, and how the systems model keyframes.more » « less
-
While large vision-language models can generate motion graphics animations from text prompts, they regularly fail to include all spatio-temporal properties described in the prompt. We introduce MoVer, a motion verification DSL based on first-order logic that can check spatio-temporal properties of a motion graphics animation. We identify a general set of such properties that people commonly use to describe animations (e.g., the direction and timing of motions, the relative positioning of objects, etc.). We implement these properties as predicates in MoVer and provide an execution engine that can apply a MoVer program to any input SVG-based motion graphics animation. We then demonstrate how MoVer can be used in an LLM-based synthesis and verification pipeline for iteratively refining motion graphics animations. Given a text prompt, our pipeline synthesizes a motion graphics animation and a corresponding MoVer program. Executing the verification program on the animation yields a report of the predicates that failed and the report can be automatically fed back to LLM to iteratively correct the animation. To evaluate our pipeline, we build a synthetic dataset of 5600 text prompts paired with ground truth MoVer verification programs. We find that while our LLM-based pipeline is able to automatically generate a correct motion graphics animation for 58.8% of the test prompts without any iteration, this number raises to 93.6% with up to 50 correction iterations. Our code and dataset are at https://mover-dsl.github.io.more » « less
-
Data science tools can help elucidate trends from clickstreams and other interactions generated by students actively using interactive textbooks. Specifically, data generated when using animations, which are multi-step visuals with text captions, will be presented in this work. Each animation step divides content into appropriate chunks, and so aligns with tenets of cognitive load theory. Both the quantity and timing of students’ clicks record provide large data sets when examining students across hundreds of animations and multiple cohorts. Specifically, an interactive textbook for a chemical engineering course in Material and Energy Balances will be examined and build upon data presented previously. While most of the previous data focused on very high reading completion rates (>99% median) compared to traditional textbooks (20-50%), a deeper examination of how long students take when watching animations will be explored. With over 140 unique animations and tens of thousands of completed views over five cohorts, a spectral clustering algorithm applied to students’ animation view times distinguished several types of animation watching behavior as well as monitor changes in this animation watching behavior over the course of a semester. After examining different numbers of clusters, two or three clusters in each chapter captured the animation usage. These clusters usually correspond to a group of students who watched animations at 1x speed (longer), another group who watched at 2x speed (shorter), and a third group, when present, who watched irregularly, including skipping animations. Overall, more students belonged to the belonged to the cluster with longer view times, with 63% of students aggregated over all cohorts and chapters compared to 35% of students in the cluster with shorter view times. The remaining 2% of students belonged to the irregular cluster, which was present in less than one quarter of the chapters. Many students stayed in the same cluster between chapters, while a smaller fraction switched between the longer and shorter clusters.more » « less
-
Generative recommendation (GR) is an emerging paradigm that tokenizes items into discrete tokens and learns to autoregressively generate the next tokens as predictions. While this token-generation paradigm is expected to surpass traditional transductive methods, potentially generating new items directly based on semantics, we empirically show that GR models predominantly generate items seen during training and struggle to recommend unseen items. In this paper, we propose SpecGR, a plug-and-play framework that enables GR models to recommend new items in an inductive setting. SpecGR uses a drafter model with inductive capability to propose candidate items, which may include both existing items and new items. The GR model then acts as a verifier, accepting or rejecting candidates while retaining its strong ranking capabilities. We further introduce the guided re-drafting technique to make the proposed candidates more aligned with the outputs of generative recommendation models, improving verification efficiency. We consider two variants for drafting: (1) using an auxiliary drafter model for better flexibility, or (2) leveraging the GR model’s own encoder for parameterefficient self-drafting. Extensive experiments on three realworld datasets demonstrate that SpecGR exhibits both strong inductive recommendation ability and the best overall performance among the compared methods.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/VIS49827.2021.9623291
