Data‐driven approaches to information access

This paper summarizes three lines of research that are motivated by the practical problem of helping users find information from external data sources, most notably computers. The application areas include information retrieval, text categorization, and question answering. A common theme in these applications is that practical information access problems can be solved by analyzing the statistical properties of words in large volumes of real world texts. The same statistical properties constrain human performance, thus we believe that solutions to practical information access problems can shed light on human knowledge representation and reasoning.     

‘He's got his father's bias’: Parental influence on weight bias in young children

Our aim was to explore the role of parents in the transmission of stereotypical body size attitudes and awareness of weight loss strategies to preschool children. Participants were 279 3‐year‐old children and their parents, who provided data at baseline and 1 year later. Parents completed self‐report body size attitude and dieting measures. Child weight bias and awareness of weight loss strategies were assessed through interview. Over time, negative associations with large bodies and awareness of weight loss strategies increased. Fathers’ attitudes prospectively predicted boys’ weight bias and awareness of weight loss strategies. Among girls, parental attitudes were less predictive. Findings confirm the importance of fathers in the development of boys’ body attitudes and inform prevention programmes.     

Is I Guaranteed to Refer?

Abstract: One claim about I , regularly made and almost universally endorsed, is that uses of the term are logically guaranteed to refer successfully (if they refer at all). The claim is only rarely formulated perspicuously or argued for. Such obscurity helps disguise the fact that those who profess to advance the claim actually turn out to support not a logical guarantee at all but merely high security through fortunate coincidence. This is not surprising. For we have no good reason to accept the claim – granted, any use of I is apt to refer successfully; but that can be explained by pragmatic features of its use. And we have some reason to reject the claim – it is notoriously difficult to see how genuine reference and guaranteed success do not exclude each other when considered as properties of the logic of any term.     

Delayed-matching-to-sample by marsh tits and great tits

The ability of two species of tits to remember the location and/or features of an object was tested in a delayed-matching-to-sample procedure. Three values of retention interval between presentation of the sample stimulus and the choice--30 sec, 5 min, and 15 min--were used. Both species performed at above-chance level at all retention intervals, and there was no significant decline in accuracy with increasing interval. A pool of 100 stimulus objects was used, but the results of control trials indicated that the birds responded primarily to location rather than stimulus features of the object itseif. Although the food-storing marsh tit tended to perform at a higher level than the non-storing great tit, the only significant difference between the species was in the first 50 trials of the first treatment, when the birds were acquiring the task. The results are discussed in relation to the hypothesized special memory capacity of food-storing birds.