Semantic Networks
Complex Sequential Question Answering: Towards Learning to Converse Over Linked Question Answer Pairs with a Knowledge Graph
Saha, Amrita (IBM Research AI) | Pahuja, Vardaan (University of Montreal) | Khapra, Mitesh M. (IIT Madras) | Sankaranarayanan, Karthik (IBM Research AI) | Chandar, Sarath (University of Montreal)
While conversing with chatbots, humans typically tend to ask many questions, a significant portion of which can be answered by referring to large-scale knowledge graphs (KG). While Question Answering (QA) and dialog systems have been studied independently, there is a need to study them closely to evaluate such real-world scenarios faced by bots involving both these tasks. Towards this end, we introduce the task of Complex Sequential QA which combines the two tasks of (i) answering factual questions through complex inferencing over a realistic-sized KG of millions of entities, and (ii) learning to converse through a series of coherently linked QA pairs. Through a labor intensive semi-automatic process, involving in-house and crowdsourced workers, we created a dataset containing around 200K dialogs with a total of 1.6M turns. Further, unlike existing large scale QA datasets which contain simple questions that can be answered from a single tuple, the questions in our dialogs require a larger subgraph of the KG. Specifically, our dataset has questions which require logical, quantitative, and comparative reasoning as well as their combinations. This calls for models which can: (i) parse complex natural language questions, (ii) use conversation context to resolve coreferences and ellipsis in utterances, (iii) ask for clarifications for ambiguous queries, and finally (iv) retrieve relevant subgraphs of the KG to answer such questions. However, our experiments with a combination of state of the art dialog and QA models show that they clearly do not achieve the above objectives and are inadequate for dealing with such complex real world settings. We believe that this new dataset coupled with the limitations of existing models as reported in this paper should encourage further research in Complex Sequential QA.
Convolutional 2D Knowledge Graph Embeddings
Dettmers, Tim (Universitร della Svizzera italiana) | Minervini, Pasquale (University College London) | Stenetorp, Pontus (University College London) | Riedel, Sebastian (University College London)
Link prediction for knowledge graphs is the task of predicting missing relationships between entities. Previous work on link prediction has focused on shallow, fast models which can scale to large knowledge graphs. However, these models learn less expressive features than deep, multi-layer models โ which potentially limits performance. In this work we introduce ConvE, a multi-layer convolutional network model for link prediction, and report state-of-the-art results for several established datasets. We also show that the model is highly parameter efficient, yielding the same performance as DistMult and R-GCN with 8x and 17x fewer parameters. Analysis of our model suggests that it is particularly effective at modelling nodes with high indegree โ which are common in highly-connected, complex knowledge graphs such as Freebase and YAGO3. In addition, it has been noted that the WN18 and FB15k datasets suffer from test set leakage, due to inverse relations from the training set being present in the test set โ however, the extent of this issue has so far not been quantified. We find this problem to be severe: a simple rule-based model can achieve state-of-the-art results on both WN18 and FB15k. To ensure that models are evaluated on datasets where simply exploiting inverse relations cannot yield competitive results, we investigate and validate several commonly used datasets โ deriving robust variants where necessary. We then perform experiments on these robust datasets for our own and several previously proposed models, and find that ConvE achieves state-of-the-art Mean Reciprocal Rank across all datasets.
TorusE: Knowledge Graph Embedding on a Lie Group
Ebisu, Takuma (SOKENDAI (The Graduate University for Advanced Studies)) | Ichise, Ryutaro (National Institute of Informatics)
Knowledge graphs are useful for many artificial intelligence (AI) tasks. However, knowledge graphs often have missing facts. To populate the graphs, knowledge graph embedding models have been developed. Knowledge graph embedding models map entities and relations in a knowledge graph to a vector space and predict unknown triples by scoring candidate triples. TransE is the first translation-based method and it is well known because of its simplicity and efficiency for knowledge graph completion. It employs the principle that the differences between entity embeddings represent their relations. The principle seems very simple, but it can effectively capture the rules of a knowledge graph. However, TransE has a problem with its regularization. TransE forces entity embeddings to be on a sphere in the embedding vector space. This regularization warps the embeddings and makes it difficult for them to fulfill the abovementioned principle. The regularization also affects adversely the accuracies of the link predictions. On the other hand, regularization is important because entity embeddings diverge by negative sampling without it. This paper proposes a novel embedding model, TorusE, to solve the regularization problem. The principle of TransE can be defined on any Lie group. A torus, which is one of the compact Lie groups, can be chosen for the embedding space to avoid regularization. To the best of our knowledge, TorusE is the first model that embeds objects on other than a real or complex vector space, and this paper is the first to formally discuss the problem of regularization of TransE. Our approach outperforms other state-of-the-art approaches such as TransE, DistMult and ComplEx on a standard link prediction task. We show that TorusE is scalable to large-size knowledge graphs and is faster than the original TransE.
Incorporating Literals into Knowledge Graph Embeddings
Kristiadi, Agustinus, Khan, Mohammad Asif, Lukovnikov, Denis, Lehmann, Jens, Fischer, Asja
Knowledge graphs, on top of entities and their relationships, contain another important element: literals. Literals encode interesting properties (e.g. the height) of entities that are not captured by links between entities alone. Most of the existing work on embedding (or latent feature) based knowledge graph modeling focuses mainly on the relations between entities. In this work, we study the effect of incorporating literal information into existing knowledge graph models. Our approach, which we name LiteralE, is an extension that can be plugged into existing latent feature methods. LiteralE merges entity embeddings with their literal information using a learnable, parametrized function, such as a simple linear or nonlinear transformation, or a multilayer neural network. We extend several popular embedding models using LiteralE and evaluate the performance on the task of link prediction. Despite its simplicity, LiteralE proves to be an effective way to incorporate literal information into existing embedding based models, improving their performance on different standard datasets, which we augmented with their literals and provide as testbed for further research.
Kojien dictionary definition for 'LGBT' criticized for inaccuracy by advocates
The publisher of Kojien, the nation's most authoritative dictionary, has drawn complaints from advocates for sexual minorities for incorrectly defining the term, "LGBT," in its latest edition released Friday. "Lesbian," "gay" and "bisexual" are terms used to describe sexual orientations, while "transgender" is used to "describe people whose gender identity does not match the sex or gender they were identified as having at birth." But the seventh edition of Kojien failed to separate the meaning of "lesbian," "gay" and "bisexual," from "transgender," defining the meaning of "LGBT" collectively as "people whose sexual orientations are different from the majority." Following its release, many LGBT advocates took to Twitter and Facebook to point out the mistake, urging the publisher to make a correction. Iwanami Shoten, the publisher, admitted the inaccuracy, saying the explanation of the term was "insufficient."
How Not to Get Lost in 2018 with Knowledge Graphs: Map, Graph, Go! - DATAVERSITY
Click to learn more about author Thomas Frisendal. Losing your way is easy. Much of Data Modeling in the search, analytics and reporting spaces have been focused on the fabulous five W-words. We have been throwing technologies at this for quite some years now: Plain old Data Modeling, Semantics, "hyperindexes", ontologies, topic maps, Data Warehouses, Operational Data Stores, multidimensional OLAP, mapping-intensive ETL, key/value pairs, Big Data and now also Data Catalogs, Data Lakes, and Knowledge Graphs. Even alerts and exception reporting have been baked into reporting and analytics solutions for quite some time now. That must be because the "what is new?" question has a bit higher priority than the other ones as business moves on.
Efficient Parallel Translating Embedding For Knowledge Graphs
Zhang, Denghui, Li, Manling, Jia, Yantao, Wang, Yuanzhuo, Cheng, Xueqi
Knowledge graph embedding aims to embed entities and relations of knowledge graphs into low-dimensional vector spaces. Translating embedding methods regard relations as the translation from head entities to tail entities, which achieve the state-of-the-art results among knowledge graph embedding methods. However, a major limitation of these methods is the time consuming training process, which may take several days or even weeks for large knowledge graphs, and result in great difficulty in practical applications. In this paper, we propose an efficient parallel framework for translating embedding methods, called ParTrans-X, which enables the methods to be paralleled without locks by utilizing the distinguished structures of knowledge graphs. Experiments on two datasets with three typical translating embedding methods, i.e., TransE [3], TransH [17], and a more efficient variant TransE- AdaGrad [10] validate that ParTrans-X can speed up the training process by more than an order of magnitude.
Taiwan objects to Kojien dictionary's definition of the independent island state
The publisher of Kojien, the most authoritative dictionary in Japan, has been stuck between a rock and a hard place over its definition of Taiwan as a province of China, prompting a request for a correction from the self-ruled island. Since its first publication in 1955, the dictionary has become a household name. The media and other organizations often use it to get the final say on a word's meaning. The seventh edition is slated to be released next month. On Friday, Iwanami Shoten, the publisher, said Kojien's entry on Taiwan is in line with the 1972 Japan-China Joint Communique, in which Japan recognized the People's Republic of China as the sole legal government of China and "fully understands and respects" the PRC's stance that Taiwan is an inalienable part of its territory.
KSR: A Semantic Representation of Knowledge Graph within a Novel Unsupervised Paradigm
Knowledge representation is a long-history topic in AI, which is very important. A variety of models have been proposed for knowledge graph embedding, which projects symbolic entities and relations into continuous vector space. However, most related methods merely focus on the data-fitting of knowledge graph, and ignore the interpretable semantic expression. Thus, traditional embedding methods are not friendly for applications that require semantic analysis, such as question answering and entity retrieval. To this end, this paper proposes a semantic representation method for knowledge graph \textbf{(KSR)}, which imposes a two-level hierarchical generative process that globally extracts many aspects and then locally assigns a specific category in each aspect for every triple. Since both aspects and categories are semantics-relevant, the collection of categories in each aspect is treated as the semantic representation of this triple. Extensive experiments show that our model outperforms other state-of-the-art baselines substantially.