Developer discussions range from in-person hallway chats to comment chains on bug reports. Being able to identify discussions that touch on software design would be helpful in documentation and refactoring software. Design mining is the application of machine learning techniques to correctly label a given discussion artifact, such as a pull request, as pertaining (or not) to design. In this work we demonstrate a simple example of how design mining works. We first replicate an existing state-of-the-art design mining study to show how conclusion stability is poor on different artifact types and different projects. Then we introduce two techniques—augmentation and context specificity—that greatly improve the conclusion stability and cross-project relevance of design mining. Our new approach achieves AUC-ROC of 0.88 on within dataset classification and 0.84 on the cross-dataset classification task. / Graduate

In this study we investigated whether a musical prime would produce a contextspecificity effect in predictive learning. Participants were divided into six conditions of a spy-radio predictive learning task. The six conditions were comprised of a combination of three primes (i.e. visual, music, or both) and two learning phase groups (i.e. retrieve, default). The primes indicated the type of stimulus used to prime the temporal context for the test cue-outcome association. The learning phase groups indicated which temporal context would be primed. In the retrieve group, learning Phase 1 was primed; in the default group learning Phase 2 was primed. The presence of a temporal contextspecificity effect was indicated by lower test predictive judgments for the test cue X in the retrieve group and higher test predictive judgments for this cue in the default group. We hypothesized that all three types of primes would lead to a significant contextspecificity effect. Furthermore, we hypothesized that the context-specificity effect would be strongest in the both prime condition because, with the presentation of both the visual and musical primes, participants would have more information about the learning phase temporal context to inform their test predictive judgment. The results partially supported the first hypothesis as there was a significant context-specificity effect with the visual prime. However, contrary to our hypotheses, we did not obtain a context-specificity effect with the music prime or both prime. Despite the lack of a context-specificity effect in the music prime condition, a relationship between participant musical expertise and predictive judgment suggested that the music did have an effect on context-specificity in predictive learning.

Finding what is alike among the personalities of creative people has been a dream of many researchers. No single personality type has been discovered as prototypical, yet the promise of common attributes among creative people remains enticing. This study examines one of these promising characteristics - Openness to Experience, a personality factor from the Five-Factor Model. This factor has been shown to correlate positively with creativity in past studies. In the present study this relationship was partially confirmed in a sample of employees whose jobs require technical problem solving, by correlating the employees self-rated Work-specific Openness to Experience and NEO PI-R Openness with supervisory ratings of their creative work performance. The Work-specific Openness scale demonstrated a significant correlation with supervisory ratings of creativity, whereas the NEO PI-R Openness scale did not. Although none of the NEO PI-R facets were significant predictors of criterion, four Work-specific facets were significant predictors based on zero order correlations. These facets are Openness to Ideas, Fantasy, Values, and Actions. However, although individual facets of Openness were expected to differ in validity, the magnitude of their correlations with creative performance scores did not differ significantly. Convincing results showing incremental validity of the Work-specific scale over the NEO PI-R scale are also discussed.

No description available.

<p>Cognitive psychologists have long been interested in the nature of the representations that underlie human behaviours. In the domain of human memory, two main accounts have been forwarded to explain how memory is represented. The multiple memory systems account posits that different phenomenological experiences of memory are the result of separate and distinct representations. In contrast, the processing account posits that there is one memory system, and that different phenomenological experiences of memory are the result of the processes engaged in when probing memory. The explanatory power of these two accounts has been evaluated in a number of domains, where it has been shown that a processing view can accommodate many of the key findings that previously led researchers to propose a separate systems framework. The goal of this thesis was to extend this work by assessing the nature of the representations that underlie implicit sequence learning. The empirical portions of this thesis provide preliminary evidence suggesting that contextual features are incorporated into implicit sequence learning. Overall, the results reported in the empirical chapters of this thesis are consistent with a processing account, as they indicate that the episodic representations that are involved in explicit remembering also support performance in implicit memory tasks.</p> / Doctor of Philosophy (PhD)

During a research process on financial speculation in relation to the home, Carl-Oskar Linné encounters statistics and reports that tell of rising unemployment among realtors. By investigating an alleged flight from the realtor profession and asking questions of prospective, current and former realtors, he attempts to understand the relationship of home sellers to their home, the housing list, the speculation mechanism and security. What are the living conditions and how is the relationship to housing policy? This essay tells about the state of the Swedish housing market and explains what has led to the work "Speculation"/"Mäklare", a work at its core questioning market ideology. The texts also describes work methods, sources of inspiration and some earlier work. / Under en researchprocess om finansiell spekulation i förhållande till hemmet stöter Carl-Oskar Linné på statistik och rapporter som berättar om en ökande arbetslöshet bland fastighetsmäklare. Genom att undersöka en påstådd flykt från mäklaryrket och ställa frågor till blivande, nuvarande och före detta fastighetsmäklare försöker han förstå bostadssäljarens förhållande sitt hem, bostadskön, spekulationsmekanismen och tryggheten. Vilka är livsvillkoren och hur är förhållandet till bostadspolitiken? Denna uppsats beskriver tillståndet på den svenska bostadsmarknaden och vad som har lett till arbetet "Mäklare", ett verk som i grunden ifrågasätter marknadsideologin. Texten beskriver även arbetsmetoder, inspirationskällor och tidigare arbete.