The requirements for software products have increased in recent years. This is both due to more complex technology as well as more requirements from the customers. An approach to solve this issue is by using a software product line (SPL) where reusable assets are developed to produce more tailor-made products with reduced time to market. When creating reusable assets, one also wants to reuse the tests for them. In order to do that, it is important to understand what to test, and where to test. A good test strategy is thereby crucial in order to avoid testing becoming a bottleneck for efficient software development. The purpose of this thesis was to investigate how to create a good testing strategy for products in an SPL. This was done by collecting information about the current testing process in order to understand which requirements that exist and how they are addressed. The core foundation of an SPL is to utilize variation to create new products. To understand the existing variation of the product and the test cases a feature modeling and similarity analysis was done. The result from them made it possible to create a test strategy and categorization of test cases that can be used to ease the reuse of test cases for new variants of the product. The resulting test strategy presented proposes feature modeling as the basis of test creation and categorization as a tool to enable easier reuse of test cases. The results of the study indicate that using SPL theory in testing can introduce a better test strategy and test case creation process which is beneficial for the whole development process. / Kraven på mjukvaruprodukter har ökat de senaste åren. Detta beror delvis på mer komplicerad teknologi, men även på att kunderna ställer högre krav på de produkter de använder. Ett tillvägagångssätt för att lösa detta är att använda en software product line. I en software product line utvecklas återanvändningsbara delar som sedan kan kombineras och därmed skapa skräddarsydda produkter på ett mer effektivt sätt. När man skapar återanvändningsbara mjukvarudelar vill man också återanvända testerna för dem. För att göra det är det viktigt att förstå vad som ska testas och varför. En bra teststrategi är därmed avgörande för att undvika att testning blir en flaskhals för effektiv mjukvaruutveckling. Syftet med detta arbete var att undersöka hur man skapar en bra teststrategi för produkter i en software product line. Detta gjordes genom att samla in information om den nuvarande testprocessen för att förstå vilka krav som finns och hur de hanteras. En huvudaspekt i software product lineteori är att använda variation för att skapa nya produkter. För att förstå variationen i produkten och dess testfall gjordes en funktionsmodellering och likhetsanalys. Resultaten från dem gjorde det möjligt att skapa en teststrategi och en kategorisering av testfall som kan användas för att underlätta återanvändningen av testfall för nya varianter av produkten. Den resulterande teststrategin som presenteras föreslår funktionsmodellering som grund för testskapande och kategorisering som ett verktyg för att möjliggöra enklare återanvändning av testfall. Resultaten av studien tyder på att användning av software product lineteori vid testning kan resultera i en bättre teststrategi och enklare process för skapandet av nya testfall.

Product line planning aims at optimal planning of product variety. In addition, the traditional product line planning problem develops new product lines based on product attributes without considering existing product lines. However, in reality, almost all new product lines evolve from existing product lines, which leads to the product line evolution problem. Product line evolution involves trade-offs between the marketing perspective and engineering perspective. The marketing concern focuses on maximizing utility for customers; the engineering concern focuses on minimizing engineering cost. Utility represents satisfaction experienced by the customers of a product. Engineering cost is the total cost involved in the process of the development of a product line. These two goals are in conflict since the high utility requires high-end product attributes which could increase the engineering cost and vice versa. Rather than aggregating both problems as one single level optimization problem, the marketing and engineering concerns entail a non-collaborative game per se. This research investigates a game-theoretic approach to the product line evolution problem. A leader-follower joint optimization model is developed to leverage conflicting goals of marketing and engineering concerns within a coherent framework of game theoretic optimization. To solve the joint optimization model efficiently, a bi-level nested genetic algorithm is developed. A case study of smart watch product line evolution is reported to illustrate the feasibility and potential of the proposed approach.

Some firms focus on few products with broad appeal, whereas others leverage variety as an important part of their offering. This study investigates the conditions under which each of these product line length strategies is optimal in the fast-moving consumer goods industry. It does so by focusing on the relationship between product line length strategy and firm top-line performance, in the light of a discrete framework derived from the strategic fit paradigm and the strategic triangle. Results show that customer factors (i.e., variety per household and value per purchase) and competitor factors (i.e., concentration and proliferation) drive the strategic fit of a firm's product line length strategy to the market in which it competes, so that firms deploying market-fitting strategies face higher odds of market share growth than firms deploying contrarian strategies. Firms with certain company factors (i.e., innovativeness and brand equity), however, face higher odds of market share growth by challenging the market fit and deploying contrarian strategies than by deploying market-fitting strategies. This study offers a stepping stone towards addressing product line length strategy as a discrete choice on product line positioning, an innovative approach with potential to generate concrete business impact.

In software product line engineering, the construction of an architecture for a product family is still an outstanding engineering challenge. In current practice, a framework is used for configuring individual products by combining solution space artefacts into products with specified features according to a feature model. No architectures are created. In contrast, an architecture for a product family would define the architectures for all the products in the family, allowing engineers to reason at a higher level of abstraction. In this thesis, we present a component model that can be used to define architectures for product families, by incorporating explicit variation points.

<p>This is a Master Thesis at the Computer Science Program at Växjö University. It has been made at BSR in Växjö where the development of a diagnostic tool has been in the running since the beginning of 2008. Previously there was a base developed within the company which I was to use for developing the two layers that I was responsible for – ProtocolLayer and PresentationLayer. In the end it should lead to a XML-based diagnostic tool. The technologies used have been .NET with its language C# and XML. The future purpose for this tool is for it to be used within the company to receive important information about the cars when developing other BSR products.</p><p>This thesis has focused on developing the tool to be working with the car manufacturer VOLVO and its protocols. The idea is that the diagnostic tool in the future could be extended to be working with all possible car manufacturers’ protocols and to enlighten the extensibility the project has been done using the product-line approach.</p>

Research in Software Product Line Engineering (SPLE) traditionally focuses on product derivation. Prior work has explored the automated derivation of products by module composition. However, it has so far neglected propagating changes (edits) in a product back to the product line definition. A domain-specific product should be possible to update its features locally, and later these changes should be propagated back to the product line definition automatically. Otherwise, the entire product line has to be revised manually in order to make the changes permanent. Although this is the current state, it is a very error-prone process. To address these issues, we present a tool called Paan to create product lines of MS Word documents with back-propagation support. It is a diff-based tool that ignores unchanged fragments and reveals fragments that are changed, added or deleted. Paan takes a document with variation points (VPs) as input, and shreds it into building blocks called tiles. Only those tiles that are new or have changed must be updated in the tile repository. In this way, changes in composed documents can be back-propagated to their original feature module definitions. A document is synthesized by retrieving the appropriate tiles and composing them. / text

Die vorliegende Arbeit beschäftigt sich mit der Frage, wie Variabilitätsinformationen aus den Quelltext von Generatoren extrahiert werden können. Zu diesem Zweck wurde eine Klassifizierung von Variablen entwickelt, die im Vergleich zu bestehenden Ansätzen eine genauere Identifikation von Merkmalen ermöglicht. Zudem bildet die Unterteilung die Basis der Erkennung von Merkmalinteraktionen und Cross-tree-Constraints. Weiterhin wird gezeigt, wie die gewonnenen Informationen durch Merkmalmodelle dargestellt werden können. Da diese auf dem Generator-Quelltext basieren, liefern sie Erkenntnisse über den Lösungsraum der Domäne. Es wird sichtbar, aus welchen Implementierungskomponenten ein Merkmal besteht und welche Beziehungen es zwischen Merkmalen gibt. Allerdings liefert ein automatisch generiertes Merkmalmodell nur wenig Erkenntnisse über den Lösungsraum. Außerdem wurde ein Prototyp entwickelt, der eine Automatisierung des beschriebenen Extraktionsprozesses ermöglicht.

Recent advances in social media have profound technical and economic implications for innovative design. This research is motivated to investigate social network effects on product design with a focus on the interface of engineering design, viral marketing, and social computing. This dissertation envisions a new paradigm of design, called viral product design for social network effects. The research problem is formulated as identification of both an optimal set of product configurations and an optimal set of seed customers so as to maximize product adoption via online social networks through equilibrium solutions to marketing-engineering coordination. Fundamental issues are investigated and a technical framework is proposed with integrated decision-based design methods. Results of case studies demonstrate that the proposed research is able to bridge the gaps between the domains of engineering design and viral marketing by incorporating social network effects. The proposed work is geared towards new design theory and decision models by integrating peer influence of social networks, which shed light on understanding the social aspect of design. The dissertation reveals the fundamental issues underlying viral product design, including the identification of viral attributes, customer preference modeling incorporating subjective experiences, the dynamics of the diffusion mechanism of online social networks, formulation of adoption maximization, and coordination between the marketing and engineering domains. In order to tackle the fundamental issues, a technical framework of viral product design for social network effects is proposed. Accordingly, mathematical and computational models are developed within the framework to support 1) latent customer needs elicitation for viral product attributes extraction, 2) customer preference modeling and quantification for product choice decision making, 3) social network modeling for product adoption prediction, and 4) viral product design evaluation by adoption maximization. These coherent models along the technical framework lay the theoretical foundation of this research, as described below. First, in order to extract potential viral product attributes, latent customer needs elicitation is emphasized. This is because latent customer needs can delight customers unexpectedly, and thus lead to potential product adoption to a large extent. We propose to elicit latent customer needs by use case analogical reasoning from sentiment analysis of online product reviews. A case study of Kindle Fire HD tablets shows the potential and feasibility of the proposed method. The extracted product attributes and attribute levels provide the choice set of viral product attributes. Second, based on the extracted product attributes, a customer preference model based on cumulative prospect theory is presented, accommodating subjective experiences in the product choice decision making process. Moreover, a hierarchical Bayesian model with Markov chain Monte Carlo is used to estimate parameters involved in the model. Based on the case study of aircraft cabin interior design, the model parameters under different experimental conditions show systematic influence of subjective experiences in choice decision making. Furthermore, a copula structure is used to construct a holistic product utility, showing customers' overall preferences to a product. This measure is crucial to product choice decision making in the context of social networks. Third, in order to predict product adoption incorporating peer influence of social networks, a linear threshold-hurdle model is proposed. It overcomes multiple drawbacks of traditional diffusion models by modeling activation thresholds, influence probability, adoption spread, holistic utility of the product, and hurdle utility of a customer in a holistic fashion. A case study of Kindle Fire HD tablets demonstrates both the predictive power of the proposed model and interesting results about customers' adoption behavior. This model paves the way for product adoption maximization in large social networks. Fourth, in order to coordinate between marketing-engineering concerns, I formulate a bi-level game theoretic optimization model for viral product design evaluation, in which the leader maximizes product adoption, while the follower optimizes product line performance. Through social network effects in terms of viral product attributes and viral influence attributes, the expected number of product adopters and the expected shared surplus, resulting from the identified product configurations and seed customers, are proved to be larger than those obtained from existing practice of viral marketing and product line design respectively, based on the case study of Kindle Fire HD tablets. Thus, the proposed paradigm of design extends the traditional boundaries among domains of engineering design, viral marketing, and social computing.

Taiwan’s recent population is growing in a declining rate, where an aging population era has arrived and struck us. The reality that only a fixed number of potential users will continue consuming motorcycle product goods has shrunk the market to an even smaller size since each household may not require as many transportation tools as they might have used to do because most of the family size structure is becoming small. On top of that, governmental policy on pushing for construction project of Metropolitan Railway Transportation (MRT) that basically benefits the general public with inexpensive convenient rides from east to west and north to south. Such construction projects have made the public eventually rely less on scooters as a necessary and convenient transportation tool. Instead, the public perceives MRT is rather time and money saving than scooters. Nonetheless, along with the growth of GDP per capita and the increasing reliance on public transportation means, the public has begun to change their perceptions on what and how they use motorcycles. The purpose of this research is to understand and investigate the relationship of the three key motorcycle players, namely, Yamaha Motors (YMT), San Yang Motors (SYM) and Kymco Motors’ (KYMCO) brand management system and their product lines decision. Since the more diversified one’s products lines is, the more complicated the process could be to manage both parent and family brand. In addition, the decision of extending one’s products lines and the synergy effects of such extensions bringing to the product equity and the entire business are worthy of studying. Since these are likely to affect how each company invests resources in R&D and market expansion.

Thesis (Ph. D.)--University of Alabama at Birmingham, 2007. / Additional advisors: Jeff G. Gray, Marjan Mernik, Rajeev Raje, Chengcui Zhang. Description based on contents viewed Feb. 7, 2008; title from title screen. Includes bibliographical references (p. 216-230).