La co-création de produits avec les consommateurs représente un axe majeur d’innovation pour les entreprises. Cependant, la question du choix de la cible à privilégier pour une co-création efficace reste en suspens. Cette recherche souhaite apporter une réponse en confrontant deux approches opposées : le crowdsourcing vs. le ciblage de certains segments de consommateurs tels que les lead-users (LU) et les emergent-nature consumers (EN). Pour répondre à cette problématique, une étude préliminaire suivie de trois études empiriques structure cette recherche. La première étude (N = 133) répond au problème d’identification des LU dans le domaine de la grande consommation en démontrant la supériorité d’une échelle de mesure face à d’autres concurrentes. La deuxième étude (N = 456) met en avant les principales motivations qui poussent les LU et les EN à co-créer et, dans le même temps, montre que ces mêmes individus sont compétents et prêts à s’engager dans des actions de co-création avec les entreprises. La troisième étude (N = 400 et N = 800), basée sur une quasi-expérimentation, révèle que le produit développé par un petit groupe de LU est globalement plus attractif que le produit développé suivant l’approche de crowdsourcing. En outre, cette dernière étude montre que les produits développés par les LU sont plus attractifs que ceux des EN. L’ensemble de ces résultats plaide en faveur du ciblage des LU pour la co-création de nouveaux produits et offre des perspectives managériales prometteuses. Enfin, ce travail doctoral apporte, de surcroît, un éclairage théorique sur le concept de co-création et ses différentes formes. / In order to innovate, companies are increasingly relying on product co-creation with consumers. However, choosing the right target for an effective co-creation remains an unresolved issue. In this research, two contrasting approaches are compared: crowdsourcing vs. targeting specific consumers such as lead-users (LU) and emergent-nature consumers (EN). To address this problem, a preliminary study was conducted, followed by three empirical studies. The first study (N = 133) focuses on the identification of LU in the domain of consumer goods and highlights the superiority of one measurement scale among others. The second study (N = 456) throws light on the main motivations that drive LU and EN to co-create and, at the same time, shows that these two targets are competent and willing to engage in co-creation activities with firms. Based on a quasi-experimentation, the third study (N = 400 and N = 800), reveals that the products developed by a small group of LU are generally more attractive than the products developed using crowdsourcing. In addition, this final study shows that the products developed by LU are more attractive than those developed by EN. All of these results suggest that companies should target LU for the co-creation of their new products and, offer promising managerial perspectives. Finally, this doctoral work provides new theoretical insights on the concept of co-creation and its various forms.

No description available.

No description available.

Expressions are very useful in a number of applications for describing the interest of the user in particular data items. Examples of such application domains include publish/subscribe, ecommerce, web site personalization. In recent work, database techniques have been utilized for efficiently matching large number of expressions with data. These techniques include storing expressions as data in the database and then indexing these expressions to quickly identify expressions that match a given data item. In this thesis a new model for expressions is presented that allows definition of richer expressions than provided in previous work. Implementation of this expression model is then described. The implementation includes sequential search as well as an indexing approach. The thesis then presents an experimental performance study that shows the benefit of the indexing approach.

This dissertation had three specific aims: to estimate the prevalence of adherence among drug users, to determine whether drug use was directly associated with adherence, and to assess whether factors drawn from the Behavioral Model of Health Care Utilization (BMHCU) are associated with adherence. The self-reported prevalence of adherence to prescribed health care recommendations in the past 12 months among the community sample of drug users and non-drug users from similar low-income areas in the study sample ranged from 53%-74%. Non-drug users consistently had higher rates of adherence than drug users, except for adherence to female health care recommendations, yet this difference was insignificant at the bivariate level. In the multivariate analysis, when controlling for all other variables within the BMHCU, only non-injection drug use was significantly associated with adherence to mental health care recommendations. In exploring its predictive utility, the BMHCU accounted for a range of 17% to 54% of the variance for the adherence measures. Although percentages accounted for by the BMHCU were substantial the fact that very few predictor variables were significant may indicate multicollinearity and other severe limitations with the data, such as small sample size, and the conceptualization of the adherence measure. The conceptualization of adherence remains an issue in need of further delineation. Further studies are needed in order to develop appropriate measures of adherence. Qualitative studies may be needed to further understand adherence among drug users.

Thesis (M.S.)--Oregon State University, 1981. / Typescript (photocopy). Includes 2 surveys in pocket. Includes bibliographical references. Also available on the World Wide Web.

This thesis examines the experiences of women becoming Internet users and readers of online texts. Millions of women in the UK began to access websites and exchange emails for the first time in the late 1990s and beginning of the 215' century. The existence of a gender gap amongst earlier adopters revealed that there must be particular obstacles facing women as new users and readers, which the thesis sets out to investigate using a multi-disciplinary approach and a range of research methods. The data is drawn from three distinct sources. First, sixteen female and six male readers were observed and interviewed while accessing their own choice of websites. This produced data on texts-in-use in conjunction with audiotaped literacy events, taking the method pioneered in schools by the Fact and Fiction Project, in its study of gender and literacy, in a new direction. Secondly, data on adult women learning to use the Internet was gathered in the form of a reflexive account of teaching beginners' classes between 2000 and 2002 at an inner-London public access location. Thirdly, women's reactions to one specific online text, an emailed petition about women in Afghanistan, were explored through an analysis of emails and comments made in interviews. The analysis situates reading online as a socially constructed literacy practice while stressing the distinctiveness of individual literacy events. It is argued that some events have particular significance, impeding or assisting the process of developing a repertoire of new literacy practices. The obstacles women face are identified as linked to the dominance of particular online texts, gendered power relations in the home and workplace, and previous educational experiences. Opportunities to create as well as read, and feeling sufficiently in control of work, access and technology, are identified as key factors in becoming a confident online reader

We study group-summarization of probabilistic graphs that naturally arise in social networks, semistructured data, and other applications. Our proposed framework groups the nodes and edges of the graph based on a user selected set of node attributes. We present methods to compute useful graph aggregates without the need to create all of the possible graph-instances of the original probabilistic graph. Also, we present an algorithm for graph summarization based on pure relational (SQL) technology. We analyze our algorithm and practically evaluate its efficiency using an extended Epinions dataset as well as synthetic datasets. The experimental results show the scalability of our algorithm and its efficiency in producing highly compressed summary graphs in reasonable time. / Graduate

No description available.

The research conducted for this thesis was focused on supporting mobile users, developing "bring your own support" solutions in order to ensure that minimum levels of computational support are always available to the user. The first line of research developed ways to minimize the negative social consequences arising from technology use. The second line of research concentrated on garment integrated devices, their use construction, and management of smart garments. The third line of research concentrated on device deployment which are another aspect of the "bring your own support" approach to supporting the mobile user.