While LLMs now enable social media users to create content classifiers easily through natural language, automatic prompt optimization techniques are often necessary to create performant classifiers. However, such techniques can fail to consider how users want to evolve their classifiers over the course of usage, including desiring to steer them in different ways during initialization and refinement. We introduce a user-centered prompt optimization technique, Promptimizer, that maintains high performance and ease-of-use but additionally (1) allows for user input into the optimization process and (2) produces final prompts that are interpretable. A lab experiment (n=16) found that users significantly preferred Promptimizer’s human-in-the-loop optimization over a fully automatic approach. We also implement Promptimizer into Puffin, a tool to support YouTube content creators in creating and maintaining personal classifiers to manage their comments. Over a 3-week deployment with 10 creators, participants successfully created diverse filters to better understand their audiences and protect their communities.

Creative writing is a deeply human craft, yet AI systems using large language models (LLMs) offer the automation of significant parts of the writing process.So why do some creative writers choose to use AI? Through interviews and observed writing sessions with 18 creative writers who already use AI regularly in their writing practice, we find that creative writers are intentional about how they incorporate AI, making many deliberate decisions about when and how to engage AI based on their core values, such as authenticity and craftsmanship.We characterize the interplay between writers' values, their fluid relationships with AI, and specific integration strategies-ultimately enabling writers to create new AI workflows without compromising their creative values.We provide insight for writing communities, AI developers and future researchers on the importance of supporting transparency of these emerging writing processes and rethinking what AI features can best serve writers.

Understanding user intents in information access scenarios can help us provide more relevant and personalized search results and recommendations. However, analyzing user intents is not easy, especially for emerging forms of Web search such as Artificial Intelligence (AI)-driven chat. To understand user intents from retrospective log data, we need a way to label them with meaningful categories that capture their diversity and dynamics. Existing methods rely on manual or Machine-Learned (ML) labeling, which is either expensive or inflexible for large and dynamic datasets. Large Language Models (LLMs) could generate rich and relevant concepts, descriptions, and examples for user intents using log data of user interactions. However, using LLMs to generate a user intent taxonomy and applying it for a given Information Retrieval (IR) application can be problematic for two main reasons: (1) such a taxonomy is not externally validated; and (2) there may be an undesirable feedback loop if an LLM does both these tasks without external validation. To address this, we propose a new methodology with human experts and assessors to verify the quality of the LLM-generated taxonomy. We also present an end-to-end pipeline that uses an LLM with Human-in-the-Loop (HITL) to produce, refine, and apply labels for user intent analysis in log data. We demonstrate its effectiveness by uncovering new insights into user intents from search and chat logs from the Microsoft Bing Web search engine. The novelty in this research stems from the method for generating purpose-driven user intent taxonomies with strong validation. Our approach not only helps remove methodological and practical bottlenecks from intent-focused research, but also provides a new framework for generating, validating, and applying other kinds of taxonomies in a scalable and adaptable way, with reasonable human effort.

Internet users have formed a wide array of online communities with diverse community goals and nuanced norms. However, most online platforms only offer a limited set of governance models in their software infrastructure and leave little room for customization. Consequently, technical proficiency becomes a prerequisite for online communities to build governance policies in code, excluding non-programmers from participation in designing community governance. In this paper, we present Pika, a system that empowers non-programmers to author a wide range of executable governance policies. At its core, Pika incorporates a declarative language that decomposes governance policies into modular components, thereby facilitating expressive policy authoring through a user-friendly, form-based web interface. Our user studies with 10 non-programmers and 7 programmers show that Pika can empower non-programmers to author policies approximately 2.5 times faster than programmers who author in code. We also provide insights about Pika’s expressivity in supporting diverse policies online communities want.

Abstract Existing visualization recommendation systems commonly rely on a single snapshot of a dataset to suggest visualizations to users. However, exploratory data analysis involves a series of related interactions with a dataset over time rather than one‐off analytical steps. We present Solas, a tool that tracks the history of a user's data analysis, models their interest in each column, and uses this information to provide visualization recommendations, all within the user's native analytical environment. Recommending with analysis history improves visualizations in three primary ways: task‐specific visualizations use the provenance of data to provide sensible encodings for common analysis functions, aggregated history is used to rank visualizations by our model of a user's interest in each column, and column data types are inferred based on applied operations. We present a usage scenario and a user evaluation demonstrating how leveraging analysis history improves in situ visualization recommendations on real‐world analysis tasks.

The localization system, which outputs vehicle position, velocity, and attitude, is one of the fundamental components in the autonomous driving vehicle. The global pose is not only used for the planning and control system, but also an important reference for the cloud source-based HD Map building and updating. The accuracy, availability, and reliability are key requirements for the localization system to ensure that the whole system runs smoothly and efficiently.IMU/Vehicle extrinsic calibration is one of the primary jobs that should be addressed. Due to the observability issue, the IMU/vehicle relative roll cannot be calibrated by the traditional maneuver-based calibration method. In this paper, we solve this issue with the proposed Multiple Orientation-based Vehicle/IMU Extrinsic Calibration (MOVIE-Cali) method, which is evaluated by Monte Carlo simulations and experiments.When the vehicle is cornering or making a U-turn, the sideslip of the tires will have negative influence on the localization system which uses Non-Holonomic Constraints (NHC)/Wheel speed sensor measurement in the model. We derive a sideslip angle model and propose an online slip parameter calibration and compensation method to improve the localization accuracy. The performance of proposed method has been evaluated by the vehicle tests.

Visible light communications (VLC) are gaining popularity and may provide an alternative means of communications in indoor settings. However, to date, there is very little research on the deployment or higher layer protocol design for VLC. In this paper, we first perform channel measurements using a physical layer testbed in the visible light band to understand its physical layer characteristics. Our measurements suggest that in order to increase data rates with VLC (1) the beam width of a communicating link can be shrunk, and (2) the transmission beam can be tuned to point towards the target recipient. We then perform Matlab simulations to verify that the human eye is able to accommodate the changes brought by shrinking a beam or by tuning the beam direction appropriately. As our main contribution, we then design a configuration framework for a VLC indoor local area network, which we call VICO; we leverage the above features towards achieving the highest throughput while maintaining fairness. VICO first tunes the beamwidths and pointing angles of the transmitters to configurations that provide the highest throughput for each client. It then tries to schedule transmissions while accounting for conflicts and the VLC PHY characteristics. Finally, it opportunistically tunes the idle LEDs to reinforce existing transmissions to increase throughput to the extent possible. We perform extensive simulations to demonstrate the effectiveness of VICO. We find that VICO provides as much as 5-fold increase in throughput compared to a simple scheduler that does not exploit the possible variations in beamwidth or beam-angle.

Ultraviolet (UV) communication is an attractive option for tactical networks or environmental monitoring. The underlying UV PHY layer has unique characteristics that render previously proposed higher layer protocols for RF communications inappropriate or inefficient. Neighbor discovery is an important functional component of a UV ad hoc wireless network. While there has been some work in UV PHY layers, there is very limited work in network study. In this paper, we propose a new neighbor discovery protocol for this setting; unlike prior protocols, our approach alleviates the negative effects of random access based collisions by choosing a leader that arbitrates the discovery process. Without prior knowledge of the number of nodes in the network, the approach facilitates neighbor discovery in a fast, fair and efficient manner. We perform extensive simulations with a realistic UV PHY layer and demonstrate that the approach reduces the required neighbor discovery time by as much as 90%. We also examine the impact of various system parameters that can be especially useful to UV network and system designers.

The solar blind ultraviolet (UV) scattering channel makes non-line-of-sight UV communications very attractive for military applications, particularly for communication on-the-move with low probability of detection and low probability of interception. Despite significant research effort on the UV physical layer, work on protocol design at the upper layers is quite limited. We consider a mobile ad hoc UV network, with each node equipped with a transceiver capable of transmitting in multiple directions and performing omni-directional receptions. Full-duplexing is enabled. We develop efficient neighbor discovery protocols by accounting for the unique UV physical (PHY) layer characteristics, namely varying channel qualities along different scattering directions. In addition to a list of neighbor nodes' identities, our protocols also construct and maintain a table which contains a ranked list of node pointing directions between each pair of nodes in terms of channel qualities. Our approach does not need support from the global positioning system (GPS) or temporal synchronization across nodes like many radio frequency (RF) protocols. Specifically, two algorithms are proposed with and without the need for direction synchronization. We further improve the latter by more efficiently utilizing neighbor feedback. We perform extensive simulations to evaluate our algorithms.