Distributional Similarity of Words with Different Frequencies

Distributional semantics tries to characterize the meaning of words by the contexts in which they occur. Similarity of words hence can be derived from the similarity of contexts. Contexts of a word are usually vectors of words appearing near to that word in a corpus. It was observed in previous research that similarity measures for the context vectors of two words depend on the frequency of these words. In the present paper we investigate this dependency in more detail for one similarity measure, the Jensen-Shannon divergence. We give an empirical model of this dependency and propose the deviation of the observed Jensen-Shannon divergence from the divergence expected on the basis of the frequencies of the words as an alternative similarity measure. We show that this new similarity measure is superior to both the Jensen-Shannon divergence and the cosine similarity in a task, in which pairs of words, taken from Wordnet, have to be classified as being synonyms or not.

A cross-sectional study assessing the association between online ratings and structural and quality of care measures : results from two German physician rating websites

BACKGROUND: Even though physician rating websites (PRWs) have been gaining in importance in both practice and research, little evidence is available on the association of patients' online ratings with the quality of care of physicians. It thus remains unclear whether patients should rely on these ratings when selecting a physician. The objective of this study was to measure the association between online ratings and structural and quality of care measures for 65 physician practices from the German Integrated Health Care Network "Quality and Efficiency" (QuE). METHODS: Online reviews from two German PRWs were included which covered a three-year period (2011 to 2013) and included 1179 and 991 ratings, respectively. Information for 65 QuE practices was obtained for the year 2012 and included 21 measures related to structural information (N = 6), process quality (N = 10), intermediate outcomes (N = 2), patient satisfaction (N = 1), and costs (N = 2). The Spearman rank coefficient of correlation was applied to measure the association between ratings and practice-related information. RESULTS: Patient satisfaction results from offline surveys and the patients per doctor ratio in a practice were shown to be significantly associated with online ratings on both PRWs. For one PRW, additional significant associations could be shown between online ratings and cost-related measures for medication, preventative examinations, and one diabetes type 2-related intermediate outcome measure. There again, results from the second PRW showed significant associations with the age of the physicians and the number of patients per practice, four process-related quality measures for diabetes type 2 and asthma, and one cost-related measure for medication. CONCLUSIONS: Several significant associations were found which varied between the PRWs. Patients interested in the satisfaction of other patients with a physician might select a physician on the basis of online ratings. Even though our results indicate associations with some diabetes and asthma measures, but not with coronary heart disease measures, there is still insufficient evidence to draw strong conclusions. The limited number of practices in our study may have weakened our findings.

On the effect of word frequency on distributional similarity

The dependency of word similarity in vector space models on the frequency of words has been noted in a few studies, but has received very little attention. We study the influence of word frequency in a set of 10 000 randomly selected word pairs for a number of different combinations of feature weighting schemes and similarity measures. We find that the similarity of word pairs for all methods, except for the one using singular value decomposition to reduce the dimensionality of the feature space, is determined to a large extent by the frequency of the words. In a binary classification task of pairs of synonyms and unrelated words we find that for all similarity measures the results can be improved when we correct for the frequency bias.