China
Africa
Explore chapters and articles related to this topic
Measurement of Cognitive States in Test and Evaluation
Published in Samuel G. Charlton, Thomas G. O’Brien, Handbook of Human Factors Testing and Evaluation, 2019
SA-SWORD. Another promising rating scale technique was developed from the Subjective Workload Dominance (SWORD) methodology by changing the instructions to participants (Vidulich & Hughes, 1991). The SA-SWORD data-collection procedure is essentially a variant of Saaty’s Analytic Hierarchy Process (Saaty, 1980) and requires participants to complete a series of pair-wise comparisons between tasks, events, or interfaces, identifying which one of each pair is associated with the highest workload. When used to assess situation awareness, the method is known as SA-SWORD. Participants rate the display or interface in each pair that produced the best situation awareness, using a 19-point scale (corresponding to scores from 2 to 9 for the stronger member of each pair, ½ to 19 for the weaker member, and a score of 1 when the members are equal). Ratings are calculated by constructing a matrix of the individual scores, calculating the geometric mean for each row, and normalizing the means accordingly (Vidulich, Ward, & Schueren, 1991). SA-SWORD has been very limited, but the initial report indicates that it was sensitive, consistent, and showed a markedly different pattern than workload ratings collected in the same scenarios (Vidulich & Hughes, 1991).
Evaluation of drinking water quality using heavy metal pollution indexing models in an agrarian, non-industrialised area of South-East Nigeria
Published in Journal of Environmental Science and Health, Part A, 2020
Onyenmechi J. Afonne, Jane U. Chukwuka, Emeka C. Ifediba
The HPI is used to determine the aggregate influence of individual heavy metal on the overall quality of water. It is a model that rates the composite influence of individual heavy metal on the overall quality of water and developed in two steps. First, by establishing a rating scale for each selected parameter giving weightage and second, by selecting the pollution parameter on which the index is to be based. The rating system is an arbitrary value between zero to one and its selection is dependent upon the importance of individual quality considerations in a comparative way, or it can be assessed by making values inversely proportional to the recommended standard for the corresponding parameter.[13,20] In the present model the unit weightage (Wi) was taken as a value inversely proportional to the recommended standard (Si) of the corresponding parameter (Table 2). The NSDWQ[7] was used for the calculation of Wi, while the metals Al, Cd, Cr, Cu, Ni and Pb were measured for the model index application.
Addressing cold start in recommender systems with neural networks: a literature survey
Published in International Journal of Computers and Applications, 2023
Research evaluated has used real-world datasets, which have numerical rating on a scale 1–5, for their experiments. Other than that rating, other ratings are Continuous ratings [41] – the ratings are given on a continuous scale and reflect how much the subject matter is liked or not liked.Interval-based ratings [41] – in interval-based ratings, ratings are frequently determined using a 5- or 7-point scale, for example, numerical integer values between 1 and 5, between −2 and 2, or between 1 and 7. An essential presumption is that the ratings are normally equally spaced apart and that the numerical values unambiguously describe the distances between them [41].Ordinal ratings [41] – like interval-based ratings, with the exception that they can also be used to rate categorical data, for example, responses such as ‘strongly disagree,’ ‘disagree,’ ‘neutral,’ ‘agree,’ and ‘strongly agree.’Binary ratings [41] – there are just two alternatives available in binary ratings, which stand for positive or negative feedback.Unary ratings [41] – these systems let the user indicate a positive preference for a product, but there is no way to indicate a negative choice, such as the case in several real-world situations, e.g. Facebook ‘like’ button [41].
Proposed RCFS-CARS Framework with Noise Detection and Correction
Published in Applied Artificial Intelligence, 2019
Veer Sain Dixit, Parul Jain, Shalini Gupta
Natural noise is the one which is involuntarily entered by users that could be because of inconsistent user behavior or erroneous or careless preference selection and the various rating collection processes that are employed (O’ Mahony et al. 2006; Li et al. 2013). Several approaches are introduced to deal with the natural noise problem. Said et al. (2012), O’ Mahony, Hurley and Silvestre (2006), and Toledo, Mota, and Martinez (2015) said to ignore noisy ratings, while others (Amatriain, Tintarev and Oliver 2009) asked to correct them by user re-rating. It remains a challenge to detect and correct those inconsistent ratings that might produce biased recommendations.