Learning to "Read Between the Lines" using Bayesian Logic Programs
Raghavan, Sindhu and Mooney, Raymond and Ku, Hyeonseo

Article Structure

Abstract

Most information extraction (IE) systems identify facts that are explicitly stated in text.

Introduction

The task of information extraction (IE) involves automatic extraction of typed entities and relations from unstructured text.

Related Work

Several previous projects (Nahm and Mooney, 2000; Carlson et al., 2010; Schoenmackers et al., 2010; Doppa et al., 2010; Sorower et al., 2011) have mined inference rules from data automatically extracted from text by an IE system.

Bayesian Logic Programs

Bayesian logic programs (BLPs) (Kersting and De Raedt, 2007; Kersting and Raedt, 2008) can be considered as templates for constructing directed graphical models (Bayes nets).

Learning BLPs to Infer Implicit Facts

4.1 Learning Rules from Extracted Data

Experimental Evaluation

5.1 Data

Results and Discussion

6.1 Comparison to Baselines

Future Work

A primary goal for future research is developing an online structure learner for BLPs that can directly learn probabilistic first-order rules from uncertain training data.

Conclusions

We have introduced a novel approach using Bayesian Logic Programs to learn to infer implicit information from facts extracted from natural language text.

Topics

ILP

Appears in 7 sentences as: ILP (7)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. tween the lines.” We present an experimental evaluation of our resulting system on a realistic test corpus from DARPA’s Machine Reading project, and demonstrate improved performance compared to a purely logical approach based on Inductive Logic Programming ( ILP ) (Lavrac and DZeroski, 1994), and an alternative SRL approach based on Markov Logic Networks (MLNs) (Domingos and Lowd, 2009).
    Page 2, “Introduction”
  2. (2010) modify an ILP system similar to FOIL (Quinlan, 1990) to learn rules with probabilistic conclusions.
    Page 2, “Related Work”
  3. (2010) use FARMER (Nijssen and Kok, 2003), an existing ILP system, to learn first-order rules.
    Page 2, “Related Work”
  4. We then learn first-order rules from these extracted facts using LIME (Mc-creath and Sharma, 1998), an ILP system designed for noisy training data.
    Page 3, “Learning BLPs to Infer Implicit Facts”
  5. Typically, an ILP system takes a set of positive and negative instances for a target relation, along with a background knowledge base (in our case, other facts extracted from the same document) from which the positive instances are potentially inferable.
    Page 3, “Learning BLPs to Infer Implicit Facts”
  6. We initially tried using the popular ALEPH ILP system (Srinivasan, 2001), but it did not produce useful rules, probably due to the high level of noise in our training data.
    Page 3, “Learning BLPs to Infer Implicit Facts”
  7. However, in contrast to MLNs, BLPs that use first-order rules that are learned by an off-the-shelf ILP system and given simple intuitive hand-coded weights, are able to provide fairly high-precision inferences that augment the output of an IE system and allow it to effectively “read between the lines.”
    Page 8, “Results and Discussion”

See all papers in Proc. ACL 2012 that mention ILP.

See all papers in Proc. ACL that mention ILP.

Back to top.

natural language

Appears in 6 sentences as: natural language (6)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. However, in natural language , some facts are implicit, and identifying them requires “reading between the lines”.
    Page 1, “Abstract”
  2. It involves learning uncertain commonsense knowledge (in the form of probabilistic first-order rules) from natural language text by mining a large corpus of automatically extracted facts.
    Page 1, “Abstract”
  3. To the best of our knowledge, this is the first paper that employs BLPs for inferring implicit information from natural language text.
    Page 2, “Introduction”
  4. We demonstrate that it is possible to learn the structure and the parameters of BLPs automatically using only noisy extractions from natural language text, which we then use to infer additional facts from text.
    Page 2, “Introduction”
  5. In our task, the supervised training data consists of facts that are extracted from the natural language text.
    Page 4, “Learning BLPs to Infer Implicit Facts”
  6. We have introduced a novel approach using Bayesian Logic Programs to learn to infer implicit information from facts extracted from natural language text.
    Page 8, “Conclusions”

See all papers in Proc. ACL 2012 that mention natural language.

See all papers in Proc. ACL that mention natural language.

Back to top.

entity types

Appears in 4 sentences as: entity types (4)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. Unlike our system and others (Carlson et al., 2010; Doppa et al., 2010; Sorower et al., 2011) that use a predefined ontology, they automatically identify a set of entity types and relations using “open IE.” They use HOLMES (Schoenmackers et al., 2008), an inference engine based on MLNs (Domingos and Lowd, 2009) (an SRL approach that combines first-order logic and Markov networks) to infer additional facts.
    Page 2, “Related Work”
  2. It consists of 57 entity types and 79 relations.
    Page 4, “Experimental Evaluation”
  3. The entity types include Agent, PhysicalThing, Event, TimeLocation, Gender, and Group, each with several subtypes.
    Page 4, “Experimental Evaluation”
  4. LIME learned several rules that had only entity types in their bodies.
    Page 4, “Experimental Evaluation”

See all papers in Proc. ACL 2012 that mention entity types.

See all papers in Proc. ACL that mention entity types.

Back to top.

knowledge base

Appears in 4 sentences as: knowledge base (4)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. Since manually developing such a knowledge base is difficult and arduous, an effective alternative is to automatically learn such rules by mining a substantial database of facts that an IE system has already automatically extracted from a large corpus of text (Nahm and Mooney, 2000).
    Page 1, “Introduction”
  2. Given a knowledge base as a BLP, standard logical inference (SLD resolution) is used to automatically construct a Bayes net for a given problem.
    Page 3, “Bayesian Logic Programs”
  3. Typically, an ILP system takes a set of positive and negative instances for a target relation, along with a background knowledge base (in our case, other facts extracted from the same document) from which the positive instances are potentially inferable.
    Page 3, “Learning BLPs to Infer Implicit Facts”
  4. The final knowledge base included all unique rules learned from any subset.
    Page 4, “Experimental Evaluation”

See all papers in Proc. ACL 2012 that mention knowledge base.

See all papers in Proc. ACL that mention knowledge base.

Back to top.

UA

Appears in 4 sentences as: UA (3) “UA” (1)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. The “unadjusted” ( UA ) score, does not correct for errors made by the extractor.
    Page 5, “Experimental Evaluation”
  2. UA AD Precision 29.73 (443/1490) 35.24 (443/1257)
    Page 6, “Experimental Evaluation”
  3. “UA” and “AD” refer to the unadjusted and adjusted scores respectively
    Page 6, “Experimental Evaluation”
  4. Table 2 gives the unadjusted ( UA ) and adjusted (AD) precision for logical deduction.
    Page 6, “Results and Discussion”

See all papers in Proc. ACL 2012 that mention UA.

See all papers in Proc. ACL that mention UA.

Back to top.

graphical model

Appears in 3 sentences as: graphical model (2) graphical models (1)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. Unlike BLPs, this approach does not use a well-founded probabilistic graphical model to compute coherent probabilities for inferred facts.
    Page 2, “Related Work”
  2. However, MLNs include all possible type-consistent groundings of the rules in the corresponding Markov net, which, for larger datasets, can result in an intractably large graphical model .
    Page 2, “Related Work”
  3. Bayesian logic programs (BLPs) (Kersting and De Raedt, 2007; Kersting and Raedt, 2008) can be considered as templates for constructing directed graphical models (Bayes nets).
    Page 3, “Bayesian Logic Programs”

See all papers in Proc. ACL 2012 that mention graphical model.

See all papers in Proc. ACL that mention graphical model.

Back to top.

manually evaluate

Appears in 3 sentences as: manual evaluation (1) manually evaluate (2)
In Learning to "Read Between the Lines" using Bayesian Logic Programs
  1. (2010) used a human judge to manually evaluate the quality of the learned rules before using them to infer additional facts.
    Page 2, “Related Work”
  2. The lack of ground truth annotation for inferred facts prevents an automated evaluation, so we resorted to a manual evaluation .
    Page 5, “Experimental Evaluation”
  3. Since it is not feasible to manually evaluate all the inferences made by the MLN, we calculated precision using only the top 1000 inferences.
    Page 6, “Results and Discussion”

See all papers in Proc. ACL 2012 that mention manually evaluate.

See all papers in Proc. ACL that mention manually evaluate.

Back to top.

PDF available from www.aclweb.org.
Send us an email