Semantics Of Answers Research Articles

Relation-Aware Heterogeneous Graph Network for Learning Intermodal Semantics in Textbook Question Answering.

Textbook question answering (TQA) task aims to infer answers for given questions from a multimodal context, including text and diagrams. The existing studies have aggregated intramodal semantics extracted from a single modality but have yet to capture the intermodal semantics between different modalities. A major challenge in learning intermodal semantics is maintaining lossless intramodal semantics while bridging the gap of semantics caused by heterogeneity. In this article, we propose an intermodal relation-aware heterogeneous graph network (IMR-HGN) to extract the intermodal semantics for TQA, which aggregates different modalities while learning features rather than representing them independently. First, we design a multidomain consistent representation (MDCR) to eliminate semantic gaps by capturing intermodal features while maintaining lossless intramodal semantics in multidomains. Furthermore, we present neighbor-based relation inpainting (NRI) to reduce semantic ambiguity via repairing fuzzy relations with correlations of relations. Finally, we propose hierarchical multisemantics aggregation (HMSA) to guarantee the completeness of semantics by aggregating features of nodes and relations with a reconstruction network (RN). Experimental results show that IMR-HGN could extract the intermodal semantics of answers, achieving a 2.16% improvement on the validation set of the TQA dataset and a 3.04% increase on the test set of the AI2D dataset.

The syntax and semantics of negative questions and answers in Korean and English

Differences in Korean and English negative polarity questions (NPQs) are revealed by the interpretation of simple yes-no answers to them. Yes-no answers to NPQs have seemingly unpredictable interpretations (Claus et al. 2017, Holmberg 2013, Kim 2017, Krifka 2017, Kramer & Rawlins 2009, Ladd 1981, Sudo 2013). However, one clearly observable fact is that yes-no answers to English and Korean NPQs can have opposite interpretations. This study: (i) compares the interpretation of positive and negative polarity questions (PPQs and NPQs) in English and Korean; (ii) examines the structure of negation in each language and its interaction with NPQs; and (iii) reports on an online experiment which gathered native speaker interpretations of NPQs in each language under context-free conditions.

Response Systems: The Syntax and Semantics of Fragment Answers and Response Particles

This article critically reviews the main research issues raised in the study of response systems in natural languages by addressing the syntax and semantics of fragment answers and yes/no response particles. Fragment answers include replies that do not have a sentential form, whereas response particles consist solely of an affirmative or a negative adverb. While the main research question in the syntax of fragments and response particles has been whether these contain more syntactic structure than what is actually pronounced, the key issues in the study of their semantics are question–answer congruence, the anaphoric potential of response particles, and the meaning of fragments in relation to positive and negative questions. In connection to these issues, this review suggests some interesting avenues for further research: ( a) providing an analysis of particles other than yes/no, ( b) choosing between echoic versus nonechoic forms as answers to polar questions, and ( c) deciding whether some non-lexically-based or nonverbal responses are systematically used in combination with polar particles to express (dis)agreement.

On the Codd semantics of SQL nulls

Theoretical models used in database research often have subtle differences with those occurring in practice. One particular mismatch that is usually neglected concerns the use of marked nulls to represent missing values in theoretical models of incompleteness, while in an SQL database these are all denoted by the same syntactic ▪ object. It is commonly argued that results obtained in the model with marked nulls carry over to SQL, because SQL nulls can be interpreted as Codd nulls, which are simply marked nulls that do not repeat. This argument, however, does not take into account that even simple queries may produce answers where distinct occurrences of ▪ do in fact denote the same unknown value. For such queries, interpreting SQL nulls as Codd nulls would incorrectly change the semantics of query answers. To use results about Codd nulls for real-life SQL queries, we need to understand which queries preserve the Codd interpretation of SQL nulls. We show, however, that the class of relational algebra queries preserving Codd interpretation is not recursively enumerable, which necessitates looking for sufficient conditions for such preservation. Those can be obtained by exploiting the information provided by NOT NULL constraints on the database schema. We devise mild syntactic restrictions on queries that guarantee preservation, do not limit the full expressiveness of queries on databases without nulls, and can be checked efficiently.

Is ‘No’ a Force-Indicator? Yes, Sooner or Later!

This paper discusses the philosophical and logical motivations for rejectivism, primarily by considering a dialogical approach to logic, which is formalized in a Question–Answer Semantics. We develop a generalised account of rejectivism through close consideration of Mark Textor’s arguments against rejectivism that the negative expression ‘No’ is never used as an act of rejection and is equivalent with a negative sentence. In doing so, we also shed light upon well-known issues regarding the supposed non-embeddability and non-iterability of force indicators.

SQL’s Three-Valued Logic and Certain Answers

The goal of the article is to bridge the difference between theoretical and practical approaches to answering queries over databases with nulls. Theoretical research has long ago identified the notion of correctness of query answering over incomplete data: one needs to find certain answers, which are true regardless of how incomplete information is interpreted. This serves as the notion of correctness of query answering, but carries a huge complexity tag. In practice, on the other hand, query answering must be very efficient, and to achieve this, SQL uses three-valued logic for evaluating queries on databases with nulls. Due to the complexity mismatch, the two approaches cannot coincide, but perhaps they are related in some way. For instance, does SQL always produce answers we can be certain about? This is not so: SQL’s and certain answers semantics could be totally unrelated. We show, however, that a slight modification of the three-valued semantics for relational calculus queries can provide the required certainty guarantees. The key point of the new scheme is to fully utilize the three-valued semantics, and classify answers not into certain or noncertain, as was done before, but rather into certainly true, certainly false, or unknown. This yields relatively small changes to the evaluation procedure, which we consider at the level of both declarative (relational calculus) and procedural (relational algebra) queries. These new evaluation procedures give us certainty guarantees even for queries returning tuples with null values.

Answering Counting Aggregate Queries over Ontologies of the DL-Lite Family

One of the main applications of description logics is the ontology-based data access model, which requires algorithms for query answering over ontologies. In fact, some description logics, like those in the DL-Lite family, are designed so that simple queries, such as conjunctive queries, are efficiently computable. In this paper we study counting aggregate queries over ontologies, i.e. queries which use aggregate functions COUNT and COUNT DISTINCT. We propose an intuitive semantics for certain answers for these queries, which conforms to the open world assumption. We compare our semantics with other approaches that have been proposed in different contexts. We establish data and combined computational complexity for the problems of answering counting aggregate queries over ontologies for several variants of DL-Lite.

Integrating domain heterogeneous data sources using decomposition aggregation queries

The decomposition aggregation query (DAQ) we introduce in this paper extends semantic integration queries by allowing query translation to create aggregate queries based on the DAQ's novel three role structure. We describe the application of DAQs in integrating domain heterogeneous data sources, the new semantics of DAQ answers and the query translation algorithm called “aggregation rewriting”.A central problem of optimizing DAQ processing requires determining the data sources towards which the DAQ is translated. Our source selection algorithm has cover-finding and partitioning steps which are optimized to 1. lower the processing overhead while speeding up query answering and 2. eliminate duplicates with minimal overhead. We establish connections between source selection optimizations and classic NP-hard optimizations and resolve the optimization problems with efficient solvers. We empirically study both the DAQ query translation and the source selection algorithms using real-world and synthetic data sets; the results show satisfying scalability both in size of aggregations and data sources for the query translation algorithms and the source selection algorithms save a good amount of computational resources.

Achieving Data Privacy through Secrecy Views and Null-Based Virtual Updates

We may want to keep sensitive information in a relational database hidden from a user or group thereof. We characterize sensitive data as the extensions of secrecy views. The database, before returning the answers to a query posed by a restricted user, is updated to make the secrecy views empty or a single tuple with null values. Then, a query about any of those views returns no meaningful information. Since the database is not supposed to be physically changed for this purpose, the updates are only virtual, and also minimal. Minimality makes sure that query answers, while being privacy preserving, are also maximally informative. The virtual updates are based on null values as used in the SQL standard. We provide the semantics of secrecy views, virtual updates, and secret answers (SAs) to queries. The different instances resulting from the virtually updates are specified as the models of a logic program with stable model semantics, which becomes the basis for computation of the SAs.

CONTEXTUAL LANGUAGE MODELS FOR RANKING ANSWERS TO NATURAL LANGUAGE DEFINITION QUESTIONS

Question–answering systems make good use of knowledge bases (KBs, e.g.,Wikipedia) for responding to definition queries. Typically, systems extract relevant facts from articles regarding the question across KBs, and then they are projected into the candidate answers. However, studies have shown that the performance of this kind of method suddenly drops, whenever KBs supply narrow coverage. This work describes a new approach to deal with this problem by constructingcontext modelsfor scoring candidate answers, which are, more precisely, statisticaln‐gram language models inferred fromlexicalized dependency pathsextracted fromWikipediaabstracts. Unlike state‐of‐the‐art approaches,context modelsare created by capturing the semantics of candidate answers (e.g., “novel,”“singer,”“coach,” and “city”). This work is extended by investigating the impact oncontext modelsof extra linguistic knowledge such as part‐of‐speech tagging and named‐entity recognition. Results showed the effectiveness ofcontext modelsasn‐gram lexicalized dependency paths and promising context indicators for the presence of definitions in natural language texts.

Returning Clustered Results for Keyword Search on XML Documents

Keyword search is an effective paradigm for information discovery and has been introduced recently to query XML documents. In this paper, we address the problem of returning clustered results for keyword search on XML documents. We first propose a novel semantics for answers to an XML keyword query. The core of the semantics is the conceptually related relationship between keyword matches, which is based on the conceptual relationship between nodes in XML trees. Then, we propose a new clustering methodology for XML search results, which clusters results according to the way they match the given query. Two approaches to implement the methodology are discussed. The first approach is a conventional one which does clustering after search results are retrieved; the second one clusters search results actively, which has characteristics of clustering on the fly. The generated clusters are then organized into a cluster hierarchy with different granularities to enable users locate the results of interest easily and precisely. Experimental results demonstrate the meaningfulness of the proposed semantics as well as the efficiency of the proposed methods.

An integrated framework for the diagnosis and correction of rule-based programs

We present a generic scheme for the declarative debugging of programs that are written in rewriting-based languages that are equipped with narrowing. Our aim is to provide an integrated development environment in which it is possible to debug a program and then correct it automatically. Our methodology is based on the combination (in a single framework) of a semantics-based diagnoser that identifies those parts of the code that contain errors and an inductive learner that tries to repair them, once the bugs have been located in the program. We develop our methodology in several steps. First, we associate with our programs a semantics that is based on a (continuous) immediate consequence operator, T R , which models the answers computed by narrowing and is parametric w.r.t. the evaluation strategy, which can be eager or lazy. Then, we show that, given the intended specification of a program R , it is possible to check the correctness of R by a single step of T R . In order to develop an effective debugging method, we approximate the computed answers semantics of R and derive a finitely terminating bottom-up abstract diagnosis method, which can be used statically. Finally, a bug-correction program synthesis methodology attempts to correct the erroneous components of the wrong code. We propose a hybrid, top-down (unfolding-based) as well as bottom-up (induction-based), correction approach that is driven by a set of evidence examples which are automatically produced as an outcome by the diagnoser. The resulting program is proven to be correct and complete w.r.t. the considered example sets. Our debugging framework does not require the user to provide error symptoms in advance or to answer difficult questions concerning program correctness. An implementation of our debugging system has been undertaken which demonstrates the workability of our approach.

Towards an algebraic theory of information integration

Information integration systems provide uniform interfaces to varieties of heterogeneous information sources. For query answering in such systems, the current generation of query answering algorithms in local-as-view (source-centric) information integration systems all produce what has been thought of as “the best obtainable” answer, given the circumstances that the source-centric approach introduces incomplete information into the virtual global relations. However, this “best obtainable” answer does not include all information that can be extracted from the sources because it does not allow partial information. Neither does the “best obtainable” answer allow for composition of queries, meaning that querying a result of a previous query will not be equivalentto the composition of the two queries. In this paper, we provide a foundation for information integration, based on the algebraic theory of incomplete information. Our framework allows us to define the semantics of partial facts and introduce the notion of the exact answer—that is the answer that includes partial facts. We show that querying under the exact answer semantics is compositional. We also present two methods for actually computing the exact answer. The first method is tableau-based, and it is a generalization of the “inverse-rules” approach. The second, much more efficient method, is a generalization of the rewriting approach, and it is based on partial containment mappings introduced in the paper.

Operational and goal-independent denotational semantics for Prolog with cut

In this paper we propose an operational and a denotational semantics for Prolog. We deal with the control rules of Prolog and the cut operator. Our denotational semantics provides a goal-independent semantics. This means that the behaviour of a goal in a program is defined as the evaluation of the goal in the denotation (semantics) of the program. We show how our denotational semantics can be specialised into a computed answer semantics and into a call pattern semantics. Our work provides a basis for a precise abstract interpretation of Prolog programs.

Conjunctive partial deduction: foundations, control, algorithms, and experiments

Partial deduction in the Lloyd–Shepherdson framework cannot achieve certain optimisations which are possible by unfold/fold transformations. We introduce conjunctive partial deduction, an extension of partial deduction accommodating such optimisations, e.g., tupling and deforestation. We first present a framework for conjunctive partial deduction, extending the Lloyd–Shepherdson framework by considering conjunctions of atoms (instead of individual atoms) for specialisation and renaming. Correctness results are given for the framework with respect to computed answer semantics, least Herbrand model semantics, and finite failure semantics. Maintaining the well-known distinction between local and global control, we describe a basic algorithm for conjunctive partial deduction, and refine it into a concrete algorithm for which we prove termination. The problem of finding suitable renamings which remove redundant arguments turns out to be important, so we give an independent technique for this. A fully automatic implementation has been undertaken, which always terminates. Differences between the abstract semantics and Prolog's left-to-right execution motivate deviations from the abstract technique in the actual implementation, which we discuss. The implementation has been tested on an extensive set of benchmarks which demonstrate that conjunctive partial deduction indeed pays off, surpassing conventional partial deduction on a range of small to medium-size programs, while remaining manageable in an automatic and terminating system.