Authors have first represented methodological issues of telephone surveys. The trends show that telephone surveying remains one the most popular mode to collect data, because it is simple to obtain probability samples of general population, with acceptable response and coverage rates. The problem is that the number of households without fixed phone is increasing, but these household have at least one mobile phone. But in the case of mobile phone surveys the most inconvenient are increased costs. According to that mixed mode surveys are one of the most promising answers of modern survey methodology to the key problems. The authors have first showed how to analytically develop the product equation and compute the exact optimal mixture for a given setting. The second example illustrates how to perform the postsurvey evaluation for mixed-mode surveys, where costs and errors are treated simultaneously.
COBISS.SI-ID: 31854429
Active involvement in discussion-based communities is nowadays a firm part of people's online activities. The measurement of communication ties and networks between contributors to such domains is thus becoming a relevant research question in social sciences. However, especially in web forums, very often almost no direct relational information exists that would indicate the presence of communication ties among contributors. In contrast with the reply-to structure of Usenet newsgroup or mailing list conversations that contain explicit relational information created by the contributors, some web forums only enable participants to add new posts to threads or to quote preceding posts in threads. When discussions emerge, it is difficult to identify who is replying to whom. Drawing on the social network studies dealing with the conversational patterns in Usenet and web forums, this paper presents an alternative approach to identifying the ties between authors of posts. Several assumptions are discussed, and different measures are developed and empirically evaluated. The findings provide a starting point for the development of a standardized methodology for studying social networks in online communities where only limited direct information about communication ties is available.
COBISS.SI-ID: 30132573
The authors of the chapter have addressed only the statistical and cost related problems of the optimal dual frame mixture and corresponding costs. First they represented dual frame designs and three ways of multiple frame research literature. In addition we are informed about the analytic solution of the mixture parameter. They have studied how to mix the mobile and fixed telephone subsamples in dual frame surveys. The authors had found that if the true population value is in the middle, it is optimal to take 50% of each subsample to form the population estimate. But if there is no difference among them, only the fixed telephone should be used, because it is cheaper.
COBISS.SI-ID: 31170141
We analyse the structure of imprecise Markov chains and study their convergence by means of accessibility relations. We first identify the sets of states, so-called minimal permanent classes, that are the minimal sets capable of containing and preserving the whole probability mass of the chain. These classes generalise the essential classes known from the classical theory. We then define a class of extremal imprecise invariant distributions and show that they are uniquely determined by the values of the upper probability on minimal permanent classes. Moreover, we give conditions for unique convergence to these extremal invariant distributions.
COBISS.SI-ID: 32156509
High-dimensional data arise naturally in many domains, and have regularly presented a great challenge for traditional data-mining techniques, both in terms of effectiveness and efficiency. Clustering becomes difficult due to the increasing sparsity of such data, as well as the increasing difficulty in distinguishing distances between data points. In this paper we take a novel perspective on the problem of clustering high-dimensional data. Instead of attempting to avoid the curse of dimensionality by observing a lower-dimensional feature subspace, we embrace dimensionality by taking advantage of some inherently high-dimensional phenomena. More specifically, we show that hubness, i.e., the tendency of high-dimensional data to contain points (hubs) that frequently occur in k-nearest neighbor lists of other points, can be successfully exploited in clustering. We validate our hypothesis by proposing several hubness-based clustering algorithms and testing them on high-dimensional data. Experimental results demonstrate good performance of our algorithms in multiple settings, particularly in the presence of large quantities of noise.
COBISS.SI-ID: 26713639