Collaborative filtering systems have been developed to manage information overload and improve discourse in online communities. In such systems, users rank content provided by other users on the validity or usefulness within their particular context. The goal is that "good" content will rise to prominence and "bad" content will fade into obscurity. These filtering mechanisms are not well-understood and have known weaknesses. For example, they depend on the presence of a large crowd to rate content, but such a crowd may not be present. Additionally, the community's decisions determine which voices will reach a large audience and which will be silenced, but it is not known if these decisions represent "the wisdom of crowds" or a "censoring mob." Our approach uses statistical machine learning to predict community ratings. By extracting features that replicate the community's verdict, we can better understand collaborative filtering, improve the way the community uses the ratings of their members, and design agents that augment community decision-making. Slashdot is an example of such a community where peers will rate each others' comments based on their relevance to the post. This work extracts a wide variety of features from the Slashdot metadata and posts' linguistic contents to identify features that can predict the community rating. We find that author reputation, use of pronouns, and author sentiment are salient. We achieve 76% accuracy predicting community ratings as good, neutral, or bad.

The success of autonomous vehicles has made path planning in real, physically grounded environments an increasingly important problem. In environments where speed matters and vehicles must maneuver around obstructions, such as autonomous car navigation in hostile environments, the speed with which real vehicles can traverse a path is often dependent on the sharpness of the corners on the path as well as the length of path edges. We present an algorithm that incorporates the use of the turn angle through path nodes as a limiting factor for vehicle speed. Vehicle speed is then used in a time-weighting calculation for each edge. This allows the path planning algorithm to choose potentially longer paths, with less turns in order to minimize path traversal time. Results simulated in the Breve environment show that travel time can be reduced over the solution obtained using the Anytime D* Algorithm by approximately 10% for a vehicle that is speed limited based on turn rate.

The projects that we have developed for UC Berkeley’s introductory artificial intelligence (AI) course teach foundational concepts using the classic video game Pac-Man. There are four project topics: state-space search, multi-agent search, probabilistic inference, and reinforcement learning. Each project requires students to implement general-purpose AI algorithms and then to inject domain knowledge about the Pac- Man environment using search heuristics, evaluation functions, and feature functions. We have found that the Pac-Man theme adds consistency to the course, as well as tapping in to students’ excitement about video games.

Reinforcement learning (RL) algorithms have the ability to learn optimal policies for control problems by exploring a domain's state space. Unfortunately, for most problems the size of the state space is too great for RL technologies to fully explore in order to find good policies. State abstraction is one way of reducing the size and complexity of a domain's state space in order to enable RL. In this paper we introduce a new approach for automatically deriving state abstractions called Evolutionary Tile Coding that uses a genetic algorithm for deriving effective tile codings. We provide an empirical analysis of the new algorithm comparing it to another adaptive tile coding method as well as fixed tile coding. Our results show that our approach is able to automatically derive effective state abstractions for two RL benchmark problems. Additionally, we present an intriguing result that shows the classical mountain car problem's state space can be reduced to just two states and still preserve the discovery of an optimal policy.

Past research has shown that when tree-structured background knowledge is available, it can be exploited to increase the efficiency of classification learning. When this kind of background knowledge is available, the problem becomes one of compositional classification. Of course, if the background knowledge contains errors, the quality of the learned hypothesis will suffer. In this paper we study the effect of faulty knowledge engineering on compositional classification learning. We present and analyze empirical results that show the impact on the quality of compositional classification learning as the quality of knowledge engineering is degraded.

This paper investigates how readily an unsolvable constraint satisfaction problem can be reformulated so that it becomes solvable. We investigate small changes in the definitions of the problemís constraints, changes that alter neither the structure of its constraint graph nor the tightness of its constraints. Our results show that structured and unstructured problems respond differently to such changes, as do easy and difficult problems taken from the same problem class. Several plausible explanations for this behavior are discussed.

We present an AI called Visuo that guesses quantitative visuospatial magnitudes (e.g., heights, lengths) given adjective-noun pairs as input (e.g., “big hat”). It uses a database of tagged images as memory and infers unexperienced magnitudes by analogy with semantically-related concepts in memory. We show that transferring width-height ratios from a semantically-related concept yields significantly lower error rates than using dissimilar concepts when predicting the width-height ratios of novel inputs.

In this paper, we introduce Bayesian Abductive Logic Programs (BALPs), a new formalism that integrates Bayesian Logic Programs (BLPs) and Abductive Logic Programming (ALP) for abductive reasoning. Like BLPs, BALPs also combine first-order logic and Bayesian networks. However, unlike BLPs that use logical deduction to construct Bayes nets, BALPs employ logical abduction. As a result, BALPs are more suited for solving problems like plan/activity recognition and diagnosis that require abductive reasoning. First, we present the necessary enhancements to BLPs in order to support logical abduction. Next, we apply BALPs to the task of plan recognition and demonstrate its efficacy on two data sets. We also compare the performance of BALPs with several existing approaches for abduction.

Learning the relational structure of a domain is a fundamental problem in statistical relational learning. The deep transfer algorithm of Davis and Domingos attempts to improve structure learning in Markov logic networks by harnessing the power of transfer learning, using the second-order structural regularities of a source domain to bias the structure search process in a target domain. We propose that the clique-scoring process which discovers these second-order regularities constitutes a novel standalone method for learning the structure of Markov logic networks, and that this fact, rather than the transfer of structural knowledge across domains, accounts for much of the performance benefit observed via the deep transfer process. This claim is supported by experiments in which we find that clique scoring within a single domain often produces results equaling or surpassing the performance of deep transfer incorporating external knowledge, and also by explicit algorithmic similarities between deep transfer and other structure learning techniques.

Most of the existing weight-learning algorithms for Markov Logic Networks (MLNs) use batch training which becomes computationally expensive and even infeasible for very large datasets since the training examples may not fit in main memory. To overcome this problem, previous work has used online learning algorithms to learn weights for MLNs. However, this prior work has only applied existing online algorithms, and there is no comprehensive study of online weight learning for MLNs. In this paper, we derive new online algorithms for structured prediction using the primal-dual framework, apply them to learn weights for MLNs, and compare against existing online algorithms on two large, real-world datasets. The experimental results show that the new algorithms achieve better accuracy than existing methods.