Exploiting inference from semantic annotations for information retrieval

The increasing amount of information that is annotated against standardised semantic resources offers opportunities to incorporate sophisticated levels of reasoning, or inference, into the retrieval process. In this position paper, we reflect on the need to incorporate semantic inference into retrieval (in particular for medical information retrieval) as well as previous attempts that have been made so far with mixed success. Medical information retrieval is a fertile ground for testing inference mechanisms to augment retrieval. The medical domain offers a plethora of carefully curated, structured, semantic resources, along with well established entity extraction and linking tools, and search topics that intuitively require a number of different inferential processes (e.g., conceptual similarity, conceptual implication, etc.). We argue that integrating semantic inference in information retrieval has the potential to uncover a large amount of information that otherwise would be inaccessible; but inference is also risky and, if not used cautiously, can harm retrieval.

Context-based interaction design for smart homes

This paper presents the notion of Context-based Activity Design (CoBAD) that represents context with its dynamic changes and normative activities in an interactive system design. The development of CoBAD requires an appropriate context ontology model and inference mechanisms. The incorporation of norms and information field theory into Context State Transition Model, and the implementation of new conflict resolution strategies based on the specific situation are discussed. A demonstration of CoBAD using a human agent scenario in a smart home is also presented. Finally, a method of treating conflicting norms in multiple information fields is proposed.

Belief revision for adaptive negotiation agents

Existing negotiation agents are primitive in terms of what they can learn and how responsive they are towards the changing negotiation contexts. These weaknesses can be alleviated if an expressive representation language is used to represent negotiation contexts and a sound inference mechanism is applied to reason about the preferential changes arising in these negotiation contexts. This paper illustrates a novel adaptive negotiation agent model, which is underpinned by the well-known AGM belief revision logic. Our preliminary experiments show that the performance of the belief-based adaptive negotiation agents is promising.

Semantic Modelling for Product Line Engineering

The aim of our work is to present solutions and a methodical support for automated techniques and procedures in domain engineering, in particular for variability modeling. Our approach is based upon Semantic Modeling concepts, for which semantic description, representation patterns and inference mechanisms are defined. Thus, model-driven techniques enriched with semantics will allow flexibility and variability in representation means, reasoning power and the required analysis depth for the identification, interpretation and adaptation of artifact properties and qualities.

Neural mechanisms of belief inference during cooperative games.

Humans have the arguably unique ability to understand the mental representations of others. For success in both competitive and cooperative interactions, however, this ability must be extended to include representations of others' belief about our intentions, their model about our belief about their intentions, and so on. We developed a "stag hunt" game in which human subjects interacted with a computerized agent using different degrees of sophistication (recursive inferences) and applied an ecologically valid computational model of dynamic belief inference. We show that rostral medial prefrontal (paracingulate) cortex, a brain region consistently identified in psychological tasks requiring mentalizing, has a specific role in encoding the uncertainty of inference about the other's strategy. In contrast, dorsolateral prefrontal cortex encodes the depth of recursion of the strategy being used, an index of executive sophistication. These findings reveal putative computational representations within prefrontal cortex regions, supporting the maintenance of cooperation in complex social decision making.

Mechanisms and molecular regulation of mammalian tooth replacement

In most non-mammalian vertebrates, such as fish and reptiles, teeth are replaced continuously. However, tooth replacement in most mammals, including human, takes place only once and further renewal is apparently inhibited. It is not known how tooth replacement is genetically regulated, and little is known on the physiological mechanism and evolutionary reduction of tooth replacement in mammals. In this study I have attempted to address these questions. In a rare human condition cleidocranial dysplasia, caused by a mutation in a Runt domain transcription factor Runx2, tooth replacement is continued. Runx2 mutant mice were used to investigate the molecular mechanisms of Runx2 function. Microarray analysis from dissected embryonic day 14 Runx2 mutant and wild type dental mesenchymes revealed many downstream targets of Runx2, which were validated using in situ hybridization and tissue culture methods. Wnt signaling inhibitor Dkk1 was identified as a candidate target, and in tissue culture conditions it was shown that Dkk1 is induced by FGF4 and this induction is Runx2 dependent. These experiments demonstrated a connection between Runx2, FGF and Wnt signaling in tooth development and possibly also in tooth replacement. The role of Wnt signaling in tooth replacement was further investigated by using a transgenic mouse model where Wnt signaling mediator β-catenin is continuously stabilized in dental epithelium. This stabilization led to activated Wnt signaling and to the formation of multiple enamel knots. In vitro and transplantation experiments were performed to examine the process of extra tooth formation. We showed that new teeth were continuously generated and that new teeth form from pre-existing teeth. A morphodynamic activator-inhibitor model was used to simulate enamel knot formation. By increasing the intrinsic production rate of the activator (β-catenin), the multiple enamel knot phenotype was reproduced by computer simulations. It was thus concluded that β-catenin acts as an upstream activator of enamel knots, closely linking Wnt signaling to the regulation of tooth renewal. As mice do not normally replace teeth, we used other model animals to investigate the physiological and genetic mechanisms of tooth replacement. Sorex araneus, the common shrew was earlier reported to have non-functional tooth replacement in all antemolar tooth positions. We showed by histological and gene expression studies that there is tooth replacement only in one position, the premolar 4 and that the deciduous tooth is diminished in size and disappears during embryogenesis without becoming functional. The growth rates of deciduous and permanent premolar 4 were measured and it was shown by competence inference that the early initiation of the replacement tooth in relation to the developmental stage of the deciduous tooth led to the inhibition of deciduous tooth morphogenesis. It was concluded that the evolutionary loss of deciduous teeth may involve the early activation of replacement teeth, which in turn suppress their predecessors. Mustela putorius furo, the ferret, has a dentition that resembles that of the human as ferrets have teeth that belong to all four tooth families, and all the antemolar teeth are replaced once. To investigate the replacement mechanism, histological serial sections from different embryonic stages were analyzed. It was noticed that tooth replacement is a process which involves the growth and detachment of the dental lamina from the lingual cervical loop of the deciduous tooth. Detachment of the deciduous tooth leads to a free successional dental lamina, which grows deeper into the mesenchyme, and later buds the replacement tooth. A careful 3D analysis of serial histological sections was performed and it was shown that replacement teeth are initiated from the successional dental lamina and not from the epithelium of the deciduous tooth. The molecular regulation of tooth replacement was studied and it was shown by examination of expression patterns of candidate regulatory genes that BMP/Wnt inhibitor Sostdc1 was strongly expressed in the buccal aspect of the dental lamina, and in the intersection between the detaching deciduous tooth and the successional dental lamina, suggesting a role for Sostdc1 in the process of detachment. Shh was expressed in the enamel knot and in the inner enamel epithelium in both generations of teeth supporting the view that the morphogenesis of both generations of teeth is regulated by similar mechanisms. In summary, histological and molecular studies on different model animals and transgenic mouse models were used to investigate tooth replacement. This thesis work has significantly contributed to the knowledge on the physiological mechanisms and molecular regulation of tooth replacement and its evolutionary suppression in mammals.

The Cache Inference Problem and its Application to Content and Request Routing

In many networked applications, independent caching agents cooperate by servicing each other's miss streams, without revealing the operational details of the caching mechanisms they employ. Inference of such details could be instrumental for many other processes. For example, it could be used for optimized forwarding (or routing) of one's own miss stream (or content) to available proxy caches, or for making cache-aware resource management decisions. In this paper, we introduce the Cache Inference Problem (CIP) as that of inferring the characteristics of a caching agent, given the miss stream of that agent. While CIP is insolvable in its most general form, there are special cases of practical importance in which it is, including when the request stream follows an Independent Reference Model (IRM) with generalized power-law (GPL) demand distribution. To that end, we design two basic "litmus" tests that are able to detect LFU and LRU replacement policies, the effective size of the cache and of the object universe, and the skewness of the GPL demand for objects. Using extensive experiments under synthetic as well as real traces, we show that our methods infer such characteristics accurately and quite efficiently, and that they remain robust even when the IRM/GPL assumptions do not hold, and even when the underlying replacement policies are not "pure" LFU or LRU. We exemplify the value of our inference framework by considering example applications.

Gene Network Inference using Machine Learning and Graph Algorithms on Big Biomedical Data

Thesis (Master's)--University of Washington, 2016-03

Bayesian Inference in Exponential and Pareto populations in the presence of Outliers

This thesis Entitled Bayesian inference in Exponential and pareto populations in the presence of outliers. The main theme of the present thesis is focussed on various estimation problems using the Bayesian appraoch, falling under the general category of accommodation procedures for analysing Pareto data containing outlier. In Chapter II. the problem of estimation of parameters in the classical Pareto distribution specified by the density function. In Chapter IV. we discuss the estimation of (1.19) when the sample contain a known number of outliers under three different data generating mechanisms, viz. the exchangeable model. Chapter V the prediction of a future observation based on a random sample that contains one contaminant. Chapter VI is devoted to the study of estimation problems concerning the exponential parameters under a k-outlier model.

Missing data mechanisms and their implications on the analysis of categorical data

We review some issues related to the implications of different missing data mechanisms on statistical inference for contingency tables and consider simulation studies to compare the results obtained under such models to those where the units with missing data are disregarded. We confirm that although, in general, analyses under the correct missing at random and missing completely at random models are more efficient even for small sample sizes, there are exceptions where they may not improve the results obtained by ignoring the partially classified data. We show that under the missing not at random (MNAR) model, estimates on the boundary of the parameter space as well as lack of identifiability of the parameters of saturated models may be associated with undesirable asymptotic properties of maximum likelihood estimators and likelihood ratio tests; even in standard cases the bias of the estimators may be low only for very large samples. We also show that the probability of a boundary solution obtained under the correct MNAR model may be large even for large samples and that, consequently, we may not always conclude that a MNAR model is misspecified because the estimate is on the boundary of the parameter space.

Genome-wide association study of reproductive traits in Nellore heifers using Bayesian inference

An important goal of Zebu breeding programs is to improve reproductive performance. A major problem faced with the genetic improvement of reproductive traits is that recording the time for an animal to reach sexual maturity is costly. Another issue is that accurate estimates of breeding values are obtained only a long time after the young bulls have gone through selection. An alternative to overcome these problems is to use traits that are indicators of the reproductive efficiency of the herd and are easier to measure, such as age at first calving. Another problem is that heifers that have conceived once may fail to conceive in the next breeding season, which increases production costs. Thus, increasing heifer's rebreeding rates should improve the economic efficiency of the herd. Response to selection for these traits tends to be slow, since they have a low heritability and phenotypic information is provided only later in the life of the animal. Genome-wide association studies (GWAS) are useful to investigate the genetic mechanisms that underlie these traits by identifying the genes and metabolic pathways involved. Data from 1853 females belonging to the Agricultural Jacarezinho LTDA were used. Genotyping was performed using the BovineHD BeadChip (777 962 single nucleotide polymorphisms (SNPs)) according to the protocol of Illumina - Infinium Assay II ® Multi-Sample HiScan with the unit SQ ™ System. After quality control, 305 348 SNPs were used for GWAS. Forty-two and 19 SNPs had a Bayes factor greater than 150 for heifer rebreeding and age at first calving, respectively. All significant SNPs for age at first calving were significant for heifer rebreeding. These 42 SNPs were next or within 35 genes that were distributed over 18 chromosomes and comprised 27 protein-encoding genes, six pseudogenes and two miscellaneous noncoding RNAs. The use of Bayes factor to determine the significance of SNPs allowed us to identify two sets of 42 and 19 significant SNPs for heifer rebreeding and age at first calving, respectively, which explain 11.35 % and 6.42 % of their phenotypic variance, respectively. These SNPs provide relevant information to help elucidate which genes affect these traits.

Multilevel modelling for inference of genetic regulatory networks

Time-course experiments with microarrays are often used to study dynamic biological systems and genetic regulatory networks (GRNs) that model how genes influence each other in cell-level development of organisms. The inference for GRNs provides important insights into the fundamental biological processes such as growth and is useful in disease diagnosis and genomic drug design. Due to the experimental design, multilevel data hierarchies are often present in time-course gene expression data. Most existing methods, however, ignore the dependency of the expression measurements over time and the correlation among gene expression profiles. Such independence assumptions violate regulatory interactions and can result in overlooking certain important subject effects and lead to spurious inference for regulatory networks or mechanisms. In this paper, a multilevel mixed-effects model is adopted to incorporate data hierarchies in the analysis of time-course data, where temporal and subject effects are both assumed to be random. The method starts with the clustering of genes by fitting the mixture model within the multilevel random-effects model framework using the expectation-maximization (EM) algorithm. The network of regulatory interactions is then determined by searching for regulatory control elements (activators and inhibitors) shared by the clusters of co-expressed genes, based on a time-lagged correlation coefficients measurement. The method is applied to two real time-course datasets from the budding yeast (Saccharomyces cerevisiae) genome. It is shown that the proposed method provides clusters of cell-cycle regulated genes that are supported by existing gene function annotations, and hence enables inference on regulatory interactions for the genetic network.

ADVERTISEMENT ALLOCATION AND TRUST MECHANISMS DESIGN IN SOCIAL NETWORKS

Social network sites (SNS), such as Facebook, Google+ and Twitter, have attracted hundreds of millions of users daily since their appearance. Within SNS, users connect to each other, express their identity, disseminate information and form cooperation by interacting with their connected peers. The increasing popularity and ubiquity of SNS usage and the invaluable user behaviors and connections give birth to many applications and business models. We look into several important problems within the social network ecosystem. The first one is the SNS advertisement allocation problem. The other two are related to trust mechanisms design in social network setting, including local trust inference and global trust evaluation. In SNS advertising, we study the problem of advertisement allocation from the ad platform's angle, and discuss its differences with the advertising model in the search engine setting. By leveraging the connection between social networks and hyperbolic geometry, we propose to solve the problem via approximation using hyperbolic embedding and convex optimization. A hyperbolic embedding method, \hcm, is designed for the SNS ad allocation problem, and several components are introduced to realize the optimization formulation. We show the advantages of our new approach in solving the problem compared to the baseline integer programming (IP) formulation. In studying the problem of trust mechanisms in social networks, we consider the existence of distrust (i.e. negative trust) relationships, and differentiate between the concept of local trust and global trust in social network setting. In the problem of local trust inference, we propose a 2-D trust model. Based on the model, we develop a semiring-based trust inference framework. In global trust evaluation, we consider a general setting with conflicting opinions, and propose a consensus-based approach to solve the complex problem in signed trust networks.