Results 1 - 10
of
29
Hingeloss Markov Random Fields: Convex Inference for Structured Prediction
- In Uncertainty in Artificial Intelligence
, 2013
"... Graphical models for structured domains are powerful tools, but the computational complexities of combinatorial prediction spaces can force restrictions on models, or require approximate inference in order to be tractable. Instead of working in a combinatorial space, we use hinge-loss Markov random ..."
Abstract
-
Cited by 28 (19 self)
- Add to MetaCart
(Show Context)
Graphical models for structured domains are powerful tools, but the computational complexities of combinatorial prediction spaces can force restrictions on models, or require approximate inference in order to be tractable. Instead of working in a combinatorial space, we use hinge-loss Markov random fields (HL-MRFs), an expressive class of graphical models with log-concave density functions over continuous variables, which can represent confidences in discrete predictions. This paper demonstrates that HL-MRFs are general tools for fast and accurate structured prediction. We introduce the first inference algorithm that is both scalable and applicable to the full class of HL-MRFs, and show how to train HL-MRFs with several learning algorithms. Our experiments show that HL-MRFs match or surpass the predictive performance of state-of-the-art methods, including discrete models, in four application domains. 1
Knowledge graph identification
- In ICML workshop on Structured Learning
, 2013
"... Abstract. Large-scale information extraction systems extract massive amounts of interrelated information, but unfortunately transforming these noisy and incomplete extractions into useful knowledge or facts is a formidable challenge. We consider the setting where uncertain extractions about entities ..."
Abstract
-
Cited by 16 (10 self)
- Add to MetaCart
(Show Context)
Abstract. Large-scale information extraction systems extract massive amounts of interrelated information, but unfortunately transforming these noisy and incomplete extractions into useful knowledge or facts is a formidable challenge. We consider the setting where uncertain extractions about entities and the relations between them constitute candidate facts for inclusion in a knowledge graph. To remove noise and infer missing information in the knowledge graph, we propose knowledge graph identification: jointly reasoning about these candidate facts and their associated extraction confidences, identifying co-referent entities, and incorporating ontological information that constrains the knowledge graph. Our approach uses probabilistic soft logic (PSL), a recently-introduced probabilistic modeling framework which easily scales to millions of facts. We demonstrate the power of our method on a synthetic Linked Data corpus derived from the MusicBrainz music community and a real-world set of extractions from the NELL web-extraction project, with over 1M extractions and 70K ontological relations, showing superior results to existing methods on both datasets. 1
Probabilistic soft logic for semantic textual similarity.
- for Computational Linguistics.
, 2014
"... Abstract Probabilistic Soft Logic (PSL) is a recently developed framework for probabilistic logic. We use PSL to combine logical and distributional representations of natural-language meaning, where distributional information is represented in the form of weighted inference rules. We apply this fra ..."
Abstract
-
Cited by 8 (0 self)
- Add to MetaCart
(Show Context)
Abstract Probabilistic Soft Logic (PSL) is a recently developed framework for probabilistic logic. We use PSL to combine logical and distributional representations of natural-language meaning, where distributional information is represented in the form of weighted inference rules. We apply this framework to the task of Semantic Textual Similarity (STS) (i.e. judging the semantic similarity of naturallanguage sentences), and show that PSL gives improved results compared to a previous approach based on Markov Logic Networks (MLNs) and a purely distributional approach.
Network-Based Drug-Target Interaction Prediction with Probabilistic Soft Logic
- IEEE/ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
, 2014
"... Drug-target interaction studies are important because they can predict drugsâ unexpected therapeutic or adverse side effects. In silico predictions of potential interactions are valuable and can focus effort on in vitro experiments. We propose a prediction framework that represents the problem us ..."
Abstract
-
Cited by 6 (3 self)
- Add to MetaCart
Drug-target interaction studies are important because they can predict drugsâ unexpected therapeutic or adverse side effects. In silico predictions of potential interactions are valuable and can focus effort on in vitro experiments. We propose a prediction framework that represents the problem using a bipartite graph of drug-target interactions augmented with drug-drug and target-target similarity measures and makes predictions using probabilistic soft logic (PSL). Using probabilistic rules in PSL, we predict interactions with models based on triad and tetrad structures. We apply (blocking) techniques that make link prediction in PSL more efficient for drug-target interaction prediction. We then perform extensive experimental studies to highlight different aspects of the model and the domain, first comparing the models with different structures and then measuring the effect of the proposed blocking on the prediction performance and efficiency. We demonstrate the importance of rule weight learning in the proposed PSL model and then show that PSL can effectively make use of a variety of similarity measures. We perform an experiment to validate the importance of collective inference and using multiple similarity measures for accurate predictions in contrast to non-collective and single similarity assumptions. Finally, we illustrate that our PSL model achieves state-of-the-art performance with simple, interpretable rules and evaluate our novel predictions using online datasets.
Learning Latent Groups with Hinge-loss Markov Random Fields
"... Probabilistic models with latent variables are powerful tools that can help explain related phenomena by mediating dependencies among them. Learning in the presence of latent variables can be difficult though, because of the difficulty of marginalizing them out, or, more commonly, maximizing a lower ..."
Abstract
-
Cited by 5 (3 self)
- Add to MetaCart
(Show Context)
Probabilistic models with latent variables are powerful tools that can help explain related phenomena by mediating dependencies among them. Learning in the presence of latent variables can be difficult though, because of the difficulty of marginalizing them out, or, more commonly, maximizing a lower bound on the marginal likelihood. In this work, we show how to learn hinge-loss Markov random fields (HL-MRFs) that contain latent variables. HL-MRFs are an expressive class of undirected probabilistic graphical models for which inference of most probable explanations is a convex optimization. By incorporating latent variables into HL-MRFs, we can build models that express rich dependencies among those latent variables. We use a hard expectation-maximization algorithm to learn the parameters of such a model, leveraging fast inference for learning. In our experiments, this combination of inference and learning discovers useful groups of users and hashtags in a Twitter data set. 1.
Collective Activity Detection using Hinge-loss Markov Random Fields
"... We propose hinge-loss Markov random fields (HL-MRFs), a powerful class of continuous-valued graphical models, for high-level computer vision tasks. HL-MRFs are characterized by log-concave density functions, and are able to perform efficient, exact inference. Their templated hinge-loss potential fun ..."
Abstract
-
Cited by 3 (1 self)
- Add to MetaCart
(Show Context)
We propose hinge-loss Markov random fields (HL-MRFs), a powerful class of continuous-valued graphical models, for high-level computer vision tasks. HL-MRFs are characterized by log-concave density functions, and are able to perform efficient, exact inference. Their templated hinge-loss potential functions naturally encode soft-valued logical rules. Using the declarative modeling language probabilistic soft logic, one can easily define HL-MRFs via familiar constructs from first-order logic. We apply HL-MRFs to the task of activity detection, using principles of collective classification. Our model is simple, intuitive and interpretable. We evaluate our model on two datasets and show that it achieves significant lift over the low-level detectors. 1.
Inferring user preferences by probabilistic logical reasoning over social networks. arXiv preprint arXiv:1411.2679
, 2014
"... We propose a framework for inferring the latent attitudes or pref-erences of users by performing probabilistic first-order logical rea-soning over the social network graph. Our method answers ques-tions about Twitter users like Does this user like sushi? or Is this user a New York Knicks fan? by bui ..."
Abstract
-
Cited by 3 (0 self)
- Add to MetaCart
(Show Context)
We propose a framework for inferring the latent attitudes or pref-erences of users by performing probabilistic first-order logical rea-soning over the social network graph. Our method answers ques-tions about Twitter users like Does this user like sushi? or Is this user a New York Knicks fan? by building a probabilistic model that reasons over user attributes (the user’s location or gender) and the social network (the user’s friends and spouse), via inferences like homophily (I am more likely to like sushi if spouse or friends like sushi, I am more likely to like the Knicks if I live in New York). The algorithm uses distant supervision, semi-supervised data harvesting and vector space models to extract user attributes (e.g. spouse, edu-cation, location) and preferences (likes and dislikes) from text. The extracted propositions are then fed into a probabilistic reasoner (we investigate both Markov Logic and Probabilistic Soft Logic). Our experiments show that probabilistic logical reasoning significantly improves the performance on attribute and relation extraction, and also achieves an F-score of 0.791 at predicting a users likes or dis-likes, significantly better than two strong baselines.
Drug-Target Interaction Prediction for Drug Repurposing with Probabilistic Similarity Logic
"... The high development cost and low success rate of drug discovery from new compounds highlight the need for methods to discover alternate therapeutic effects for currently approved drugs. Computational methods can be effective in focusing efforts for such drug repurposing. In this paper, we propose a ..."
Abstract
-
Cited by 2 (2 self)
- Add to MetaCart
(Show Context)
The high development cost and low success rate of drug discovery from new compounds highlight the need for methods to discover alternate therapeutic effects for currently approved drugs. Computational methods can be effective in focusing efforts for such drug repurposing. In this paper, we propose a novel drug-target interaction prediction framework based on probabilistic similarity logic (PSL) [5]. Interaction prediction corresponds to link prediction in a bipartite network of drug-target interactions extended with a set of similarities between drugs and between targets. Using probabilistic first-order logic rules in PSL, we show how rules describing link predictions based on triads and tetrads can effectively make use of a variety of similarity measures. We learn weights for the rules based on training data, and report relative importance of each similarity for interaction prediction. We show that the learned rule weights significantly improve prediction precision. We evaluate our results on a dataset of drug-target interactions obtained from Drugbank [27] augmented with five drug-based and three target-based similarities. We integrate domain knowledge in drug-target interaction prediction and match the performance of the state-of-the-art drug-target interaction prediction systems [22] with our model using simple triad-based rules. Furthermore, we apply techniques that make link prediction in PSL more efficient for drug-target interaction prediction.
Semantic parsing using distributional semantics and probabilistic logic
- In ACL 2014 Workshop on Semantic Parsing
, 2014
"... We propose a new approach to semantic parsing that is not constrained by a fixed formal ontology and purely logical infer-ence. Instead, we use distributional se-mantics to generate only the relevant part of an on-the-fly ontology. Sentences and the on-the-fly ontology are represented in probabilist ..."
Abstract
-
Cited by 2 (0 self)
- Add to MetaCart
(Show Context)
We propose a new approach to semantic parsing that is not constrained by a fixed formal ontology and purely logical infer-ence. Instead, we use distributional se-mantics to generate only the relevant part of an on-the-fly ontology. Sentences and the on-the-fly ontology are represented in probabilistic logic. For inference, we use probabilistic logic frameworks like Markov Logic Networks (MLN) and Prob-abilistic Soft Logic (PSL). This seman-tic parsing approach is evaluated on two tasks, Textual Entitlement (RTE) and Tex-tual Similarity (STS), both accomplished using inference in probabilistic logic. Ex-periments show the potential of the ap-proach. 1
Large-Scale Knowledge Graph Identification using PSL
"... Building a web-scale knowledge graph, which captures information about entities and the relationships between them, represents a formidable challenge. While many largescale information extraction systems operate on web corpora, the candidate facts they produce are noisy and incomplete. To remove noi ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
(Show Context)
Building a web-scale knowledge graph, which captures information about entities and the relationships between them, represents a formidable challenge. While many largescale information extraction systems operate on web corpora, the candidate facts they produce are noisy and incomplete. To remove noise and infer missing information in the knowledge graph, we propose knowledge graph identification: a process of jointly reasoning about the structure of the knowledge graph, utilizing extraction confidences and leveraging ontological information. Scalability is often a challenge when building models in domains with rich structure, but we use probabilistic soft logic (PSL), a recentlyintroduced probabilistic modeling framework which easily scales to millions of facts. In practice, our method performs joint inference on a real-world dataset containing over 1M facts and 80K ontological constraints in 12 hours and produces a high-precision set of facts for inclusion into a knowledge graph. 1.
