The Alchemy System for Statistical Relational AI: User Manual

By Stanley Kok, Parag Singla, Matthew Richardson, Pedro M. Domingos, M. Poon. 2007

321 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/18f96ef4398bb3893706c8e42d944a7108dbcb91

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Entity Resolution with Markov Logic (2006, methodology): “These algorithms are publicly available in the Alchemy system, which we use in our experiments [24].”
Probabilistic programming (2014, methodology): “The Alchemy system for statistical relational AI.”
TODTLER: Two-Order-Deep Transfer Learning (2015, methodology): “We learned the weights of the formulas in the target model using Alchemy (Kok et al. 2010) and applied a pruning threshold of 0.05 on the weights of the clauses.”
Mapping and Revising Markov Logic Networks for Transfer Learning (2007, methodology): “We used the implementation of KD provided as part of the Alchemy software package (Kok et al. 2005) and implemented our new algorithms as part of the same package.”
Statistical predicate invention (2007, methodology): “MLN Structure Learning We also compare MRC to Kok and Domingos (2005) MLN structure learning algorithm (MSL, beam search version) implemented in the Alchemy package (Kok et al., 2006).”
R2: An Efficient MCMC Sampler for Probabilistic Programs (2014, background): “Over the past several years, a variety of probabilistic programming languages and inference systems have been proposed (Gilks, Thomas, and Spiegelhalter 1994; Koller, McAllester, and Pfeffer 1997; Minka et al. 2009; Pfeffer 2007a; Goodman et al. 2008; Kok et al. 2007).”
Prediction of protein beta-residue contacts by Markov logic networks with grounding-specific weights (2009, methodology): “On the other hand, the computation of ωi with a kernel machine would require time proportional to the number of cases e.g. the number of candidate partners when applied to Formula 7, while using neural networks we can compute it in constant time.”
Transfer Learning via Relational Type Matching (2015, methodology): “Similar to MLN structure learning, in MLN weight learning, we used the default settings of Alchemy to learn discriminative weights.”
Automatically refining the wikipedia infobox ontology (2008, methodology): “Using MLNs to Classify Subsumption KOG uses the open source Alchemy system [22] to implement its MLNs classifier.”
A user-guided approach to program analysis (2015, background): “There are a number of solvers [8,14,26,33,34,37] that eﬃ-ciently compile a probabilistic analysis down into a MaxSAT formula, and solve the corresponding formula to produce the desired output that satisﬁes all the hard rules and maximizes the weight of the satisﬁed soft rules.”
Learning Markov Logic Networks via Functional Gradient Boosting (2011, methodology): “We used the queryEvidence flag for Alchemy weight learning and inference.”
Sound and Efficient Inference with Probabilistic and Deterministic Dependencies (2006, methodology): “We implemented MC-SAT and simulated tempering as extensions of the Alchemy system (Kok et al. 2005), and used Alchemy’s implementation of Gibbs sampling, with ten chains.”