Abstract: Market segmentation in transportation planning is a division of a total population of travelers into groups (segments) which are relatively homogeneous with respect to certain personal characteristics (the segmentation base). It is desirable that the segments be distinct in terms of travel behavior and their reactions to changes in the travel environment such as the introduction of new transportation services. This report documents a comparison of market segmentation using six different bases. The first two segmentations represent alternative ways of dividing the population into groups which had homogeneous demographic characteristics; the second two segmentations divided the population into groups based upon such travel choice constraints as auto availability and transit service; the last two segmentations divided the population into groups based upon their responses to attitude surveys. Among these pairs of segmentations, some were multivariate, while others involved only a single variable. The six segmentations were compared with respect to five criteria judged to be important considerations in transportation planning: measurability (data availability), statistical robustness, substantiality (size and importance of resulting segments), relationship with travel behavior, and relationship with planning of service options. The comparisons showed that no single segmentation base was superior according to all criteria. On balance, however, multivariate segmentations. In addition, the segmentation based upon multivariate choice constraints satishied more of the criteria than other segmentation bases. Segmentations of the traveling population based upon attitudes were found to have certain specific uses but to be inferior to choice constraints segmentation for most planning purpose. /Author/

Abstract: The Harpy connected speech recognition system is the result of our attempt to understand the relative importance of various design choices of two earlier speech recognition systems developed at Carnegie‐Mellon University, The Hearsay‐I system (D. R. Reddy, et al., IEEE Trans. AU, 229–238 (1973); L. D. Erman, Technical Report, Computer Science Department, Carnegie‐Mellon University, (1974) and the Dragon system [J. K. Baker, Ph.D. Thesis (Carnegie‐Mellon University)]. Knowledge is represented as procedures in Hearsay‐I and as a Markov network with a‐priori transition probabilities between states in Dragon. Hearsay‐I uses a best first search while Dragon searches all the possible (acoustic syntactic) paths through the network to determine the optimal path. Hearsay‐I uses segmentation and labeling and Dragon is a segmentation‐free system. Systematic performance analysis of various design choices resulted in the Harpy system which represents knowledge as a finite state transition network but without the a‐priori transition probabilities, searches only a few “best” paths, and uses segmentation to reduce the number of state probability updates that must be done. The system achieves between 70% and 100% sentence accuracy, depending upon the task, and runs between 1.5 and 10 times real time. Complete details of design, implementation, and experimental results are given in B. P. Lowerre, Ph.D. Thesis (Computer Science Department, Carnegie‐Mellon University, 1976).

Abstract: PURPOSE:To enable dynamic change of segmentation operation during execution of a program by giving the function to modify and re-define to segment description element describing a segment.

Abstract: : In the first part of this paper, 'The Overall Design', the authors looked at their evolving understanding of computational techniques--both in analyzing the visual system of animals, and in building computer vision systems. The authors divided the computations between low-level systems and high-level systems, and sketched the interaction between the two types of system. Here, in part 2, the authors focus upon specific tasks of the low-level systems--feature extraction and segmentation, as well as their competition and cooperation. The authors continue to emphasize an integrated system design, with interaction of multiple processes resolving ambiguous and noisy data. A survey of processes which operate on a single static, but colored, image show how segmentation can proceed via boundary formation, and by formation of regions on the basis of color and texture cues. Extensive experimental data are given on the results of applying segmentation techniques.

Abstract: In the past, picture segmentation has been performed by merging small primitive regions or by recursively splitting the whole picture. This paper combines the two approaches with significant increase in processing speed while maintaining small memory requirements. The data structure is described in detail and examples of implementations are given.

Abstract: Written genealogies are used as a means of ascertaining how members of Chinese lineages regard the growth and segmentation of their own social groups. The genealogies reveal a concern, not with corporate and joint estates—the basis of segmentation in Chinese lineages according to Maurice Freedman's model—but with change in residence resulting from the formation of new settlements. The significance of this fact for understanding the indigenous view of both lineage growth and the connection between common residence and common descent is discussed.