Probabilistic Models for Scalable Knowledge Graph Construction

In the past decade, systems that extract information from millions of Internet documents have become commonplace. Knowledge graphs -- structured knowledge bases that describe entities, their attributes and the relationships between them -- are a powerful tool for understanding and organizing this vast amount of information. However, a significant obstacle to knowledge graph construction is the unreliability of the extracted information, due to noise and ambiguity in the underlying data or errors made by the extraction system and the complexity of reasoning about the dependencies between these noisy extractions. My dissertation addresses these challenges by exploiting the interdependencies between facts to improve the quality of the knowledge graph in a scalable framework. I introduce a new approach called knowledge graph identification (KGI), which resolves the entities, attributes and relationships in the knowledge graph by incorporating uncertain extractions from multiple sources, entity co-references, and ontological constraints. I define a probability distribution over possible knowledge graphs and infer the most probable knowledge graph using a combination of probabilistic and logical reasoning. Such probabilistic models are frequently dismissed due to scalability concerns, but my implementation of KGI maintains tractable performance on large problems through the use of hinge-loss Markov random fields, which have a convex inference objective. This allows the inference of large knowledge graphs using 4M facts and 20M ground constraints in 2 hours. To further scale the solution, I develop a distributed approach to the KGI problem which runs in parallel across multiple machines, reducing inference time by 90%. Finally, I extend my model to the streaming setting, where a knowledge graph is continuously updated by incorporating newly extracted facts. I devise a general approach for approximately updating inference in convex probabilistic models, and quantify the approximation error by defining and bounding inference regret for online models. Together, my work retains the attractive features of probabilistic models while providing the scalability necessary for large-scale knowledge graph construction. These models have been applied on a number of real-world knowledge graph projects, including the NELL project at Carnegie Mellon and the Google Knowledge Graph.

Du document à la donnée et retour. La fourmilière ou les Lumières

Même si le mot est ancien, la référence à la notion de document est récente dans l'Histoire, sans doute en résonance avec l'organisation de la société industrielle, sa régulation et ses valeurs. Au tournant du millénaire, le web s'est appuyé sur un renversement du circuit documentaire, jusqu'à, dans le web de données, un court-circuitage radical. S'agit-il de l'effacement d'une notion périmée au profit d'une autre ou d'un simple décalage ? Le succès du web accompagne des transformations sociales et économiques profondes. Quelles seront alors, sur nos régimes de vérité, de preuve et de transmission, les conséquences de l'éventuelle obsolescence de la notion de document ?

hyo-xtm: a set of hyper-graph operations on xml topic map toward knowledge management

Knowledge management is a critical issue for the next-generation web application, because the next-generation web is becoming a semantic web, a knowledge-intensive network. XML Topic Map (XTM), a new standard, is appearing in this field as one of the structures for the semantic web. It organizes information in a way that can be optimized for navigation. In this paper, a new set of hyper-graph operations on XTM (HyO-XTM) is proposed to manage the distributed knowledge resources.HyO-XTM is based on the XTM hyper-graph model. It is well applied upon XTM to simplify the workload of knowledge management.The application of the XTM hyper-graph operations is demonstrated by the knowledge management system of a consulting firm. HyO-XTM shows the potential to lead the knowledge management to the next-generation web.

Content aggregation on knowledge bases using graph clustering

Recently, research projects such as PADLR and SWAP have developed tools like Edutella or Bibster, which are targeted at establishing peer-to-peer knowledge management (P2PKM) systems. In such a system, it is necessary to obtain provide brief semantic descriptions of peers, so that routing algorithms or matchmaking processes can make decisions about which communities peers should belong to, or to which peers a given query should be forwarded. This paper proposes the use of graph clustering techniques on knowledge bases for that purpose. Using this clustering, we can show that our strategy requires up to 58% fewer queries than the baselines to yield full recall in a bibliographic P2PKM scenario.

Linked knowledge sources for topic classification of microposts:a semantic graph-based approach

Short text messages a.k.a Microposts (e.g. Tweets) have proven to be an effective channel for revealing information about trends and events, ranging from those related to Disaster (e.g. hurricane Sandy) to those related to Violence (e.g. Egyptian revolution). Being informed about such events as they occur could be extremely important to authorities and emergency professionals by allowing such parties to immediately respond. In this work we study the problem of topic classification (TC) of Microposts, which aims to automatically classify short messages based on the subject(s) discussed in them. The accurate TC of Microposts however is a challenging task since the limited number of tokens in a post often implies a lack of sufficient contextual information. In order to provide contextual information to Microposts, we present and evaluate several graph structures surrounding concepts present in linked knowledge sources (KSs). Traditional TC techniques enrich the content of Microposts with features extracted only from the Microposts content. In contrast our approach relies on the generation of different weighted semantic meta-graphs extracted from linked KSs. We introduce a new semantic graph, called category meta-graph. This novel meta-graph provides a more fine grained categorisation of concepts providing a set of novel semantic features. Our findings show that such category meta-graph features effectively improve the performance of a topic classifier of Microposts. Furthermore our goal is also to understand which semantic feature contributes to the performance of a topic classifier. For this reason we propose an approach for automatic estimation of accuracy loss of a topic classifier on new, unseen Microposts. We introduce and evaluate novel topic similarity measures, which capture the similarity between the KS documents and Microposts at a conceptual level, considering the enriched representation of these documents. Extensive evaluation in the context of Emergency Response (ER) and Violence Detection (VD) revealed that our approach outperforms previous approaches using single KS without linked data and Twitter data only up to 31.4% in terms of F1 measure. Our main findings indicate that the new category graph contains useful information for TC and achieves comparable results to previously used semantic graphs. Furthermore our results also indicate that the accuracy of a topic classifier can be accurately predicted using the enhanced text representation, outperforming previous approaches considering content-based similarity measures. © 2014 Elsevier B.V. All rights reserved.

Cultural Knowledge for Named Entity Disambiguation: A Graph-Based Semantic Relatedness Approach

One of the ultimate aims of Natural Language Processing is to automate the analysis of the meaning of text. A fundamental step in that direction consists in enabling effective ways to automatically link textual references to their referents, that is, real world objects. The work presented in this paper addresses the problem of attributing a sense to proper names in a given text, i.e., automatically associating words representing Named Entities with their referents. The method for Named Entity Disambiguation proposed here is based on the concept of semantic relatedness, which in this work is obtained via a graph-based model over Wikipedia. We show that, without building the traditional bag of words representation of the text, but instead only considering named entities within the text, the proposed method achieves results competitive with the state-of-the-art on two different datasets.

Graph-based concept weighting for medical information retrieval

This paper presents a graph-based method to weight medical concepts in documents for the purposes of information retrieval. Medical concepts are extracted from free-text documents using a state-of-the-art technique that maps n-grams to concepts from the SNOMED CT medical ontology. In our graph-based concept representation, concepts are vertices in a graph built from a document, edges represent associations between concepts. This representation naturally captures dependencies between concepts, an important requirement for interpreting medical text, and a feature lacking in bag-of-words representations. We apply existing graph-based term weighting methods to weight medical concepts. Using concepts rather than terms addresses vocabulary mismatch as well as encapsulates terms belonging to a single medical entity into a single concept. In addition, we further extend previous graph-based approaches by injecting domain knowledge that estimates the importance of a concept within the global medical domain. Retrieval experiments on the TREC Medical Records collection show our method outperforms both term and concept baselines. More generally, this work provides a means of integrating background knowledge contained in medical ontologies into data-driven information retrieval approaches.

Building instance knowledge network for word sense disambiguation

In this paper, a new high precision focused word sense disambiguation (WSD) approach is proposed, which not only attempts to identify the proper sense for a word but also provides the probabilistic evaluation for the identification confidence at the same time. A novel Instance Knowledge Network (IKN) is built to generate and maintain semantic knowledge at the word, type synonym set and instance levels. Related algorithms based on graph matching are developed to train IKN with probabilistic knowledge and to use IKN for probabilistic word sense disambiguation. Based on the Senseval-3 all-words task, we run extensive experiments to show the performance enhancements in different precision ranges and the rationality of probabilistic based automatic confidence evaluation of disambiguation. We combine our WSD algorithm with five best WSD algorithms in senseval-3 all words tasks. The results show that the combined algorithms all outperform the corresponding algorithms.

Knowledge discovery from tree databases using balanced optimal search

This research is a step forward in discovering knowledge from databases of complex structure like tree or graph. Several data mining algorithms are developed based on a novel representation called Balanced Optimal Search for extracting implicit, unknown and potentially useful information like patterns, similarities and various relationships from tree data, which are also proved to be advantageous in analysing big data. This thesis focuses on analysing unordered tree data, which is robust to data inconsistency, irregularity and swift information changes, hence, in the era of big data it becomes a popular and widely used data model.

Using a glow graph to represent data flow and dependency in event logs

The idea of extracting knowledge in process mining is a descendant of data mining. Both mining disciplines emphasise data flow and relations among elements in the data. Unfortunately, challenges have been encountered when working with the data flow and relations. One of the challenges is that the representation of the data flow between a pair of elements or tasks is insufficiently simplified and formulated, as it considers only a one-to-one data flow relation. In this paper, we discuss how the effectiveness of knowledge representation can be extended in both disciplines. To this end, we introduce a new representation of the data flow and dependency formulation using a flow graph. The flow graph solves the issue of the insufficiency of presenting other relation types, such as many-to-one and one-to-many relations. As an experiment, a new evaluation framework is applied to the Teleclaim process in order to show how this method can provide us with more precise results when compared with other representations.

Simulation and graph mining tools for improving gene mapping efficiency

Gene mapping is a systematic search for genes that affect observable characteristics of an organism. In this thesis we offer computational tools to improve the efficiency of (disease) gene-mapping efforts. In the first part of the thesis we propose an efficient simulation procedure for generating realistic genetical data from isolated populations. Simulated data is useful for evaluating hypothesised gene-mapping study designs and computational analysis tools. As an example of such evaluation, we demonstrate how a population-based study design can be a powerful alternative to traditional family-based designs in association-based gene-mapping projects. In the second part of the thesis we consider a prioritisation of a (typically large) set of putative disease-associated genes acquired from an initial gene-mapping analysis. Prioritisation is necessary to be able to focus on the most promising candidates. We show how to harness the current biomedical knowledge for the prioritisation task by integrating various publicly available biological databases into a weighted biological graph. We then demonstrate how to find and evaluate connections between entities, such as genes and diseases, from this unified schema by graph mining techniques. Finally, in the last part of the thesis, we define the concept of reliable subgraph and the corresponding subgraph extraction problem. Reliable subgraphs concisely describe strong and independent connections between two given vertices in a random graph, and hence they are especially useful for visualising such connections. We propose novel algorithms for extracting reliable subgraphs from large random graphs. The efficiency and scalability of the proposed graph mining methods are backed by extensive experiments on real data. While our application focus is in genetics, the concepts and algorithms can be applied to other domains as well. We demonstrate this generality by considering coauthor graphs in addition to biological graphs in the experiments.

Theory and Algorithms for Hop-Count-Based Localization with Random Geometric Graph Models of Dense Sensor Networks

Wireless sensor networks can often be viewed in terms of a uniform deployment of a large number of nodes in a region of Euclidean space. Following deployment, the nodes self-organize into a mesh topology with a key aspect being self-localization. Having obtained a mesh topology in a dense, homogeneous deployment, a frequently used approximation is to take the hop distance between nodes to be proportional to the Euclidean distance between them. In this work, we analyze this approximation through two complementary analyses. We assume that the mesh topology is a random geometric graph on the nodes; and that some nodes are designated as anchors with known locations. First, we obtain high probability bounds on the Euclidean distances of all nodes that are h hops away from a fixed anchor node. In the second analysis, we provide a heuristic argument that leads to a direct approximation for the density function of the Euclidean distance between two nodes that are separated by a hop distance h. This approximation is shown, through simulation, to very closely match the true density function. Localization algorithms that draw upon the preceding analyses are then proposed and shown to perform better than some of the well-known algorithms present in the literature. Belief-propagation-based message-passing is then used to further enhance the performance of the proposed localization algorithms. To our knowledge, this is the first usage of message-passing for hop-count-based self-localization.

Diversity in ranking via resistive graph centers

Users can rarely reveal their information need in full detail to a search engine within 1--2 words, so search engines need to "hedge their bets" and present diverse results within the precious 10 response slots. Diversity in ranking is of much recent interest. Most existing solutions estimate the marginal utility of an item given a set of items already in the response, and then use variants of greedy set cover. Others design graphs with the items as nodes and choose diverse items based on visit rates (PageRank). Here we introduce a radically new and natural formulation of diversity as finding centers in resistive graphs. Unlike in PageRank, we do not specify the edge resistances (equivalently, conductances) and ask for node visit rates. Instead, we look for a sparse set of center nodes so that the effective conductance from the center to the rest of the graph has maximum entropy. We give a cogent semantic justification for turning PageRank thus on its head. In marked deviation from prior work, our edge resistances are learnt from training data. Inference and learning are NP-hard, but we give practical solutions. In extensive experiments with subtopic retrieval, social network search, and document summarization, our approach convincingly surpasses recently-published diversity algorithms like subtopic cover, max-marginal relevance (MMR), Grasshopper, DivRank, and SVMdiv.

Multiview stereo via volumetric Graph-Cuts and occlusion robust photo-consistency.

This paper presents a volumetric formulation for the multi-view stereo problem which is amenable to a computationally tractable global optimisation using Graph-cuts. Our approach is to seek the optimal partitioning of 3D space into two regions labelled as "object" and "empty" under a cost functional consisting of the following two terms: (1) A term that forces the boundary between the two regions to pass through photo-consistent locations and (2) a ballooning term that inflates the "object" region. To take account of the effect of occlusion on the first term we use an occlusion robust photo-consistency metric based on Normalised Cross Correlation, which does not assume any geometric knowledge about the reconstructed object. The globally optimal 3D partitioning can be obtained as the minimum cut solution of a weighted graph.

SiGMa: Simple Greedy Matching for Aligning Large Knowledge Bases

The Internet has enabled the creation of a growing number of large-scale knowledge bases in a variety of domains containing complementary information. Tools for automatically aligning these knowledge bases would make it possible to unify many sources of structured knowledge and answer complex queries. However, the efficient alignment of large-scale knowledge bases still poses a considerable challenge. Here, we present Simple Greedy Matching (SiGMa), a simple algorithm for aligning knowledge bases with millions of entities and facts. SiGMa is an iterative propagation algorithm which leverages both the structural information from the relationship graph as well as flexible similarity measures between entity properties in a greedy local search, thus making it scalable. Despite its greedy nature, our experiments indicate that SiGMa can efficiently match some of the world's largest knowledge bases with high precision. We provide additional experiments on benchmark datasets which demonstrate that SiGMa can outperform state-of-the-art approaches both in accuracy and efficiency.