A translation apparatus is provided which comprises: an inputting section for inputting a source document in a natural language; a layout analyzing section for analyzing layout information including cascade information, itemization information, numbered itemization information, labeled itemization information and separator line information in the source document inputted by the inputting section and specifying a translation range on the basis of the layout information; a translation processing section for translating a source document text in the specified translation range into a second language; and an outputting section for outputting a translated text provided by the translation processing section.

Automated Planning: Theory and Practice.

We present a novel reasoning calculus for the description logic SHOIQ+--a knowledge representation formalism with applications in areas such as the SemanticWeb. Unnecessary nondeterminism and the construction of large models are two primary sources of inefficiency in the tableau-based reasoning calculi used in state-of-the-art reasoners. In order to reduce nondeterminism, we base our calculus on hypertableau and hyperresolution calculi, which we extend with a blocking condition to ensure termination. In order to reduce the size of the constructed models, we introduce anywhere pairwise blocking. We also present an improved nominal introduction rule that ensures termination in the presence of nominals, inverse roles, and number restrictions--a combination of DL constructs that has proven notoriously difficult to handle. Our implementation shows significant performance improvements over state-of-the-art reasoners on several well-known ontologies.

/pdf/hypertableau-reasoning-for-description-logics-1z3294bbzx.pdf

Hypertableau reasoning for description logics

The Semantic Web contains an enormous amount of information in the form of knowledge bases (KB). To make this information available, many question answering (QA) systems over KBs were created in the last years. Building a QA system over KBs is difficult because there are many different challenges to be solved. In order to address these challenges, QA systems generally combine techniques from natural language processing, information retrieval, machine learning and Semantic Web. The aim of this survey is to give an overview of the techniques used in current QA systems over KBs. We present the techniques used by the QA systems which were evaluated on a popular series of benchmarks: Question Answering over Linked Data. Techniques that solve the same task are first grouped together and then described. The advantages and disadvantages are discussed for each technique. This allows a direct comparison of similar techniques. Additionally, we point to techniques that are used over WebQuestions and SimpleQuestions, which are two other popular benchmarks for QA systems.

/pdf/core-techniques-of-question-answering-systems-over-knowledge-ztjne1c44r.pdf

Core techniques of question answering systems over knowledge bases: a survey

/pdf/cincinnati-children-s-hospital-medical-centerdenohai-wai-shi-1clnzi3cqo.pdf

Cincinnati Children’s Hospital Medical Centerでの海外実習を終えて

Explainable AI is not a new field. Since at least the early exploitation of C.S. Pierce’s abductive reasoning in expert systems of the 1980s, there were reasoning architectures to support an explanation function for complex AI systems, including applications in medical diagnosis, complex multi-component design, and reasoning about the real world. So explainability is at least as old as early AI, and a natural consequence of the design of AI systems. While early expert systems consisted of handcrafted knowledge bases that enabled reasoning over narrowly well-defined domains (e.g., INTERNIST, MYCIN), such systems had no learning capabilities and had only primitive uncertainty handling. But the evolution of formal reasoning architectures to incorporate principled probabilistic reasoning helped address the capture and use of uncertain knowledge.

/pdf/explainable-ai-the-new-42-280k5x29zb.pdf

Explainable AI: The new 42?

We present a domain-agnostic system for Question Answering over multiple semi-structured and possibly linked datasets without the need of a training corpus. The system is motivated by an industry use-case where Enterprise Data needs to be combined with a large body of Open Data to fulfill information needs not satisfied by prescribed application data models. Our proposed Question Answering pipeline combines existing components with novel methods to perform, in turn, linguistic analysis of a query, named entity extraction, entity/graph search, fusion and ranking of possible answers. We evaluate QuerioDALI with two open-domain benchmarks and a biomedical one over Linked Open Data sources, and show that our system produces comparable results to systems that require training data and are domain-dependent. In addition, we analyze the current challenges and shortcomings.

QuerioDALI: Question Answering Over Dynamic and Linked Knowledge Graphs

It is well known that a correspondence theory for description logics (DLs), propositional modal logics (MLs) and propositional dynamic logics (PDLs) was given in [Sch91] and that an axiomatic system for a description logic with inverse roles was presented in the same paper Schild’s paper covers what is to be discussed in this paper although his paper did not mention nominals as well as expressive roles explicitly On the other hand, reductions to eliminate converse programs (inverse roles in DLs) are known even for full PDL (ALCIreg); the original reduction from converse PDL to PDL was in [Gia96] Moreover, a direct tableaux method for converse PDL and further discussions on the elimination of converse programs were given in [GM00] Besides graded modalities, it was pointed out in [Gia96] that the previous reduction technique is also applicable to nominals This paper shares some viewpoints well made in [Gia96] The proposed mapping process here is based on the simple idea to capture possible backpropagation caused by the use of inverse roles This process consists of three steps, tagging, recording, polarisation, which are introduced below Concept expressions/formulae are assumed to be in negation norm form (NNF) For simplicity, existential restrictions and universal restrictions are called modal constraints somewhere We refer to [BCM03] for usual background knowledge on description logics (DLs)

/pdf/a-new-mapping-from-alci-to-alc-4we4b8et10.pdf

A New Mapping from ALCI to ALC.

We consider the instance checking problem over S ? I Q ( D ) $\mathcal {SHIQ}(\mathbf{D})$ knowledge bases, that is, the problem of determining if the class membership of a given object is logically implied by a knowledge base expressed in terms of the description logic (DL) dialect S ? I Q ( D ) $\mathcal {SHIQ}(\mathbf{D})$ . Such problems are inevitable in conjunctive query evaluation over such knowledge bases, or indeed for any knowledge bases that rely on an ability to capture disjunction and/or negation in an underlying DL. This includes the problem of evaluating basic graph patterns occurring in SPARQL queries over RDF graphs with the so-called OWL 2 direct semantics entailment regime, that is, where the RDF graph is an OWL 2 ontology. Our main result is a novel method for such problems that derives from an adaptation of binary absorption. We show that the method works particularly well for knowledge bases that have a very large collection of factual assertions about individual objects.

Absorption for ABoxes

Absorptions are generally employed in Description Logics (DL) reasoners in a uniform way regardless of the structure of an input knowledge base. In this paper we present an approach to encode some state-of-the-art absorption techniques into a state space planner, aiming to achieve a better solution. The planner applies appropriate operators to general axioms and produces a solution with a minimized cost to automatically organize these absorptions in a certain sequence to facilitate DL reasoning. Compared to predetermined or fixed applications of established absorptions, such a solution is more flexible and probable to absorb more general axioms into an unfoldable TBox.

/pdf/planning-of-axiom-absorption-57e82j7mmc.pdf

Planning of Axiom Absorption.

We develop a novel absorption technique for large collections of factual assertions about individual objects. These assertions are commonly accompanied by implicit background knowledge and form a knowledge base. Both the assertions and the background knowledge are expressed in a suitable language of Description Logic and queries over such knowledge bases can be expressed as assertion retrieval queries. The proposed absorption technique significantly improves the performance of such queries, in particular in cases where a large number of object features are known for the objects represented in such a knowledge base. In addition to the absorption technique we present the results of a preliminary experimental evaluation that validates the efficacy of the proposed optimization.

/pdf/assertion-absorption-in-object-queries-over-knowledge-bases-hohw1dn6ht.pdf

Jiewen Wu

Papers

QuerioDALI: Question Answering Over Dynamic and Linked Knowledge Graphs

A New Mapping from ALCI to ALC.

Absorption for ABoxes

Planning of Axiom Absorption.

Assertion absorption in object queries over knowledge bases