Active Learning for Multi-relational Data Construction

Kajino, Hiroshi; Kishimoto, Akihiro; Botea, Adi; Daly, Elizabeth; Kotoulas, Spyros

doi:10.1145/2736277.2741103

Cited by 5 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The methods in this literature can be divided into two general categories. One category is the conventional active learning framework where an algorithm selects unlabeled data points and human intelligence is employed to label this selected point [13,24,31,44,49]. The other category consists of methods that rely more on human intelligence.…”

Section: Related Workmentioning

confidence: 99%

Active Learning for Skewed Data Sets

Kazerouni,

Zhao,

Xie

et al. 2020

Preprint

View full text Add to dashboard Cite

Consider a sequential active learning problem where, at each round, an agent selects a batch of unlabeled data points, queries their labels and updates a binary classifier. While there exists a rich body of work on active learning in this general form, in this paper, we focus on problems with two distinguishing characteristics: severe class imbalance (skew) and small amounts of initial training data. Both of these problems occur with surprising frequency in many web applications. For instance, detecting offensive or sensitive content in online communities (pornography, violence, and hate-speech) is receiving enormous attention from industry as well as research communities. Such problems have both the characteristics we describe -a vast majority of content is not offensive, so the number of positive examples for such content is orders of magnitude smaller than the negative examples. Furthermore, there is usually only a small amount of initial training data available when building machine-learned models to solve such problems. To address both these issues, we propose a hybrid active learning algorithm (HAL) that balances exploiting the knowledge available through the currently labeled training examples with exploring the large amount of unlabeled data available. Through simulation results, we show that HAL makes significantly better choices for what points to label when compared to strong baselines like margin-sampling. Classifiers trained on the examples selected for labeling by HAL easily out-perform the baselines on target metrics (like area under the precision-recall curve) given the same budget for labeling examples. We believe HAL offers a simple, intuitive, and computationally tractable way to structure active learning for a wide range of machine learning applications.

show abstract

Section: Related Workmentioning

confidence: 99%

Active Learning for Skewed Data Sets

Kazerouni,

Zhao,

Xie

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Finally, RedThread is also related to measuring proximity in graphs [38], to build the k nearest neighbor structure to search over [12], and various graph extraction techniques [10,14,26] and relational data knowledge base extraction methods [16,35,36]. These problem settings however differ from ours as RedThread focuses on a local active setting targeted at finding related entities.…”

Section: Active Search Exploration and Clusteringmentioning

confidence: 99%

Active Search of Connections for Case Building and Combating Human Trafficking

Rabbany

David

Dubrawski

2018

Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery &Amp; Data Mining

View full text Add to dashboard Cite

How can we help an investigator to efficiently connect the dots and uncover the network of individuals involved in a criminal activity based on the evidence of their connections, such as visiting the same address, or transacting with the same bank account? We formulate this problem as Active Search of Connections, which finds target entities that share evidence of different types with a given lead, where their relevance to the case is queried interactively from the investigator. We present RedThread, an efficient solution for inferring related and relevant nodes while incorporating the user's feedback to guide the inference. Our experiments focus on case building for combating human trafficking, where the investigator follows leads to expose organized activities, i.e. different escort advertisements that are connected and possibly orchestrated. RedThread is a local algorithm and enables online case building when mining millions of ads posted in one of the largest classified advertising websites. The results of RedThread are interpretable, as they explain how the results are connected to the initial lead. We experimentally show that RedThread learns the importance of the different types and different pieces of evidence, while the former could be transferred between cases.

show abstract

“…Experimental settings: We compare the Thompson sampling models with Amdc models, and Prescal for passive learning. Amdc model has been proposed to achieve two different active learning goals, constructing a predictive model and maximising the valid triples in a knowledge base, with two different querying strategies [10]. Amdc-pred is a predictive model construction strategy and chooses a triple which is the most ambiguous (close to the decision boundary) at each time t. Amdc-pop is a population strategy which aims to maximise the number of valid triples in a knowledge base, choosing a triple with the highest expected value at each time.…”

Section: Knowledge Populationmentioning

confidence: 99%

“…In the original Amdc [10], Amdc-pop model obtains more valid triples than Amdc-pred, and Amdc-pred shows high ROC-AUC scores than Amdc-pop. In our experiment, however, Amdc-pop shows comparable cumulative gain to Amdcpred and even worse than Amdc-pred for the UMLS.…”

Section: Knowledge Populationmentioning

confidence: 99%

See 1 more Smart Citation

Probabilistic Knowledge Graph Construction

Kim

Xie

Ong

2016

Proceedings of the 25th ACM International on Conference on Information and Knowledge Management

View full text Add to dashboard Cite

Knowledge graph construction consists of two tasks: extracting information from external resources (knowledge population) and inferring missing information through a statistical analysis on the extracted information (knowledge completion). In many cases, insufficient external resources in the knowledge population hinder the subsequent statistical inference. The gap between these two processes can be reduced by an incremental population approach. We propose a new probabilistic knowledge graph factorisation method that benefits from the path structure of existing knowledge (e.g. syllogism) and enables a common modelling approach to be used for both incremental population and knowledge completion tasks. More specifically, the probabilistic formulation allows us to develop an incremental population algorithm that trades off exploitation-exploration. Experiments on three benchmark datasets show that the balanced exploitation-exploration helps the incremental population, and the additional path structure helps to predict missing information in knowledge completion.

show abstract

Active Learning for Multi-relational Data Construction

Cited by 5 publications

References 20 publications

Active Learning for Skewed Data Sets

Active Learning for Skewed Data Sets

Active Search of Connections for Case Building and Combating Human Trafficking

Probabilistic Knowledge Graph Construction

Contact Info

Product

Resources

About