An approach to semantic-based model discovery and selection

Abstract

Model discovery and selection is an important step in component-based simulation model development. This paper proposes an efficient model discovery approach and quantifies the degrees of semantic similarity for selection of partially matched models. Models are represented as production strings as specified by an EBNF composition grammar. Together with a novel DHT overlay network, we achieve fast discovery of syntactically similar models with discovery cost independent of the model size. Next, we rank partially matched models for selection using semantic-based model attributes and behavior. Experiments conducted on a repository with 4,000 models show that on average DHT-based model lookup using production strings takes less than one millisecond compared with two minutes using naive string comparisons. Lastly, efficient model selection is a tradeoff between query representation and the computation cost of model ranking.

Reusing simulation experiment specifications in developing models by successive composition — a case study of the Wnt/β-catenin signaling pathway

“…6,7 This can serve as a basis to structure the process of developing models via composition and identify crucial steps. Likewise, several perspectives exist to structure the process of model development based on composition, e.g., see the work by Kaputis and Ng, and Szabo and Teo, 8,9 depending on how model composition is supported by the corresponding approaches. To structure the model development process via composition (as depicted in Figure 1), we adopt the work presented by Balci and Sargent and start the life cycle with problem formulation, requirements engineering and conceptual modeling, which are merged into one phase denoted as Problem formulation, Requirements engineering and Conceptual modeling (PRC), as these are not the focus of this paper.…”

“…6,7 This can serve as a basis to structure the process of developing models via composition and identify crucial steps. Likewise, several perspectives exist to structure the process of model development based on composition, e.g., see the work by Kaputis and Ng, and Szabo and Teo, 8,9 depending on how model composition is supported by the corresponding approaches.…”

“…Accordingly, we remove the lipid rafts from the model M12, by setting nLR = 0 to reproduce the findings of our wet-lab experiments with lipid raft deficient hNPC. The two properties can be expressed based on mitl [a,b] as in Equations (10) and (9), while ses12 denotes the experiment set-up, which is similar to ses2 but nLR = 0. Simulation experiments as presented in Figure 8(c) showed that the second expectation can be met, however, not the first one, i.e., Equation (10) does NOT hold for the composed model M12, only Equation 9: In M12, the Wnt influence on b-catenin entirely relies on lipid rafts.…”

The space complexity of mirror games

“…The proof of this corollary with N + 1 in place of N follows from Lemma 3 by constructing an (Odd, Even)-town in space [N + 1] from sets B1 ∪ {N + 1}, B2 ∪ {N + 1}, ..., Bm ∪ {N + 1}. For the proof with N , refer to [Ber69,Sza11].…”

Tutorial on building M&S software based on reuse

“…More work in the M&S community deals with reusable model (components). E.g., in SimBeans (Praehofer, Sametinger, and Stritzinger 2001) the Java bean component model is facilitated, the HLA (Kuhl, Weatherly, and Dahmann 2000) allows to reuse models (computed by an extra piece of software), COMO (Röhl and Uhrmacher 2006) and the approach by Szabo and Teo (2011) focus on composing models by well defined specialized languages and mechanisms. Valentin, Verbraeck, and Sol (2003) examined the impact of using building bricks to create models, to use model libraries is common in software like MatLab/SimuLink or those based on MODELICA (Elmqvist, Mattsson, and Otter 1998).…”