|
on Big Data |
| By: | Lu, Richard (?); Chatman, Jennifer A. (?); Goldberg, Amir (Stanford University); Srivastava, Sameer B. (?) |
| Abstract: | From the schoolyard to the boardroom, the pressures of cultural assimilation pervade all walks of social life. Why are some people more successful than others at cultural adjustment? Research on organizational culture has mostly focused on value congruence as the core dimension of cultural fit. We develop a complementary conceptualization of cognitive fit--perceptual accuracy, or the degree to which a person can decipher the group's cultural code. We demonstrate that the ability to read the cultural code, rather than identification with the code, matters for contemporaneous behavioral conformity. We further show that a person*s behavior and perceptual accuracy are both influenced by observations of others* behavior, whereas value congruence is less susceptible to peer influence. Drawing on email and survey data from a mid-sized technology firm, we use the tools of computational linguistics and machine learning to develop longitudinal measures of cognitive and behavioral cultural fit. We also take advantage of a reorganization that produced quasi-exogenous shifts in employees' interlocutors to identify the causal impact of peer influence. We discuss implications of these findings for research on cultural assimilation, the interplay of structure and culture, and the pairing of surveys with digital trace data. |
| Date: | 2018–05 |
| URL: | http://d.repec.org/n?u=RePEc:ecl:stabus:3603&r=big |
| By: | Annette Kaempf-Dern; Christoph Mader |
| Abstract: | University lectures are facing higher quantities of students showing increasingly heterogeneous knowledge, interest and intellectual capabilities. To prepare those student groups for a changing, interconnected world, basic knowledge teaching is not sufficient any more. Instead, to let them acquire 21st century skills, e.g. strong communication and collaboration skills, expertise in technology, innovative and creative thinking skills, it is necessary to change the methods of teaching. Blended learning and Flipped classroom are concepts that can be combined to achieve these goals. The questions the paper addresses are: 1.) Do these concepts provide benefits when dealing with student groups that are strongly heterogeneous in knowledge, personality, culture, background? And 2.) How effective are these concepts, and which factors cause or prevent success?The research team, consisting of real estate subject lecturers and instructional learning design experts, set up an introductory class of business administration in a blended learning format and used this situation as a “real life experimental” study. During the study, 'big data' was collected and analyzed, operationalizing up to 250 students’ online behavior, their learning time slots as well as intermediary test results and direct feedback. The study was complemented by longitudinal surveys regarding the students’ motivation, workload sentiment, expectations, knowledge and other dimensions Also, a comprehensive evaluation survey was analyzed quantitatively as well as qualitatively.Survey results showed a large variance of the appreciation of the learning concepts which can be explained by several reasons. However, the workload sentiment was a major issue. Rather minor changes are sufficient to adress this problem, while the general concepts of blended learning and flipped classroom can be kept as effective learning methods. Yet, for fully assessing the learning success regarding 21st century skills, further empirical studies are necessary.The study provides recommendations of what works and what was not effective in a blended learning/flipped classroom course, e.g. how an incentive system can be designed to engage students in a constant learning process and increase interaction during the semester. The originality/value of the study is its focus on large, heterogeneous groups and its design as an experimental real-life study combining the collection of big data with quantitative and qualitative methods. |
| Keywords: | 21st Century Competences; Big data; Blended Learning; Flipped Classroom; Mixed-methods teaching concepts |
| JEL: | R3 |
| Date: | 2018–01–01 |
| URL: | http://d.repec.org/n?u=RePEc:arz:wpaper:eres2018_274&r=big |
| By: | Zhipeng Liang; Kangkang Jiang; Hao Chen; Junhao Zhu; Yanran Li |
| Abstract: | In this paper, we implement two state-of-art continuous reinforcement learning algorithms, Deep Deterministic Policy Gradient (DDPG) and Proximal Policy Optimization (PPO) in portfolio management. Both of them are widely-used in game playing and robot control. What's more, PPO has appealing theoretical propeties which is hopefully potential in portfolio management. We present the performances of them under different settings, including different learning rate, objective function, markets, feature combinations, in order to provide insights for parameter tuning, features selection and data preparation. |
| Date: | 2018–08 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:1808.09940&r=big |
| By: | Gill, Andrej; Hett, Florian; Tischer, Johannes |
| Abstract: | Improving financial conditions of individuals requires an understanding of the mechanisms through which bad financial decision-making leads to worse financial outcomes. From a theoretical point of view, a key candidate inducing mistakes in financial decision-making are so called present-biased preferences, which are one of the cornerstones of behavioral economics. According to theory, present-biased households should behave systematically different when it comes to consumption and saving decisions, as they should be more prone to spending too much and saving too little. In this policy letter we show how high frequency financial transaction data available in digitized form allows to precisely categorize individual financial-decision making to be present-biased or not. Using this categorization, we find that one out of five individuals in our sample exhibits present-bias and that this present-biased behavior is associated with a stronger use of overdrafts. As overdrafts represent a particularly expensive way of short-term borrowing, their systematic use can be interpreted as a measure of suboptimal financial-decision making. Overall, our results indicate that the combination of economic theory and Big Data is able to generate valuable insights with applications for policy makers and businesses alike. |
| Keywords: | financial decision-making,financial transaction data,present bias,time inconsistency |
| Date: | 2018 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:safewh:55&r=big |
| By: | Hillen, Judith |
| Abstract: | Web scraping is a method for extracting large amounts of data from online sources. In food price analysis, however, this data collection technique has not yet received a lot of attention. We discuss how this method can be used in food price research and identify areas of application. We find that web scraping is a promising method to collect customized, high-frequency data in real time, overcoming several limitations of currently used food price data. While today’s applications mostly focus on (online) consumer prices, the scope of applications broadens as more and more price data are published online. To better deal with the technical and legal challenges of web scraping and to exploit its scalability, joint data collection projects in the field of agricultural and food economics should be considered. |
| Date: | 2018–09–01 |
| URL: | http://d.repec.org/n?u=RePEc:ags:gewi18:275840&r=big |
| By: | Peiseler, Florian; Rasch, Alexander; Shekhar, Shiva |
| Abstract: | We analyze firms' ability to sustain collusion in a setting in which horizontally differentiated firms can price-discriminate based on private information regarding consumers' preferences. In particular, firms receive private signals which can be noisy (e.g., big data predictions). We find that there is a non-monotone relationship between signal quality and sustainability of collusion. Starting from a low level, an increase in signal precision first facilitates collusion. However, there is a turning point from which on any further increase renders collusion less sustainable. Our analysis provides important insights for competition policy. In particular, a ban on price discrimination can help to prevent collusive behavior as long as signals are sufficiently noisy. |
| Keywords: | Big Data,Collusion,Loyalty,Private Information,Third-Degree Price Discrimination |
| JEL: | L13 D43 L41 |
| Date: | 2018 |
| URL: | http://d.repec.org/n?u=RePEc:zbw:dicedp:295&r=big |
| By: | Kubitza, Christoph; Krishna, Vijesh; Urban, Kira; Qaim, Matin |
| Abstract: | The expansion of agricultural land remains one of the main drivers of deforestation in tropical regions, with severe negative environmental consequences. Stronger land property rights could possibly enable farmers to increase input intensity and productivity on the already cultivated land, thus reducing incentives to expand their farms by deforesting additional land. This hypothesis is tested with data from a panel survey of farm households in Sumatra, Indonesia, one of the hotspots of recent rainforest loss due to agricultural area expansion. The survey data are combined with satellite imageries to account for spatial patterns, such as historical forest locations. Results show that plots for which farmers hold formal land titles are cultivated more intensively than untitled plots, even after controlling for other relevant factors. Land titles also contribute to higher crop yields, hence confirming expectations. However, due to land policy restrictions, farmers located at the historic forest margins often do not hold formal titles for the land they cultivate. Without land titles, these farmers are less able to intensify and more likely to expand into the surrounding forest land to increase agricultural output. Indeed, forest closeness and past deforestation activities by households are found to be positively associated with current farm size. The findings suggest that the observed land policy restrictions are not conducive for forest conservation. In addition to improving farmer’s access to land titles for non-forest land, better recognition of customary land rights and moreeffective protection of forest land without recognized claims could be useful policy responses. |
| Keywords: | Land Economics/Use |
| Date: | 2017–08–28 |
| URL: | http://d.repec.org/n?u=RePEc:ags:eaae17:261166&r=big |
| By: | Jae-uk Choo (Chung-Ang University) |
| Abstract: | William Gibson?s Neuromancer introduces diverse types of transhumans and posthumans like cyborg, robot, A.I. These show how the future community will develop with symptoms of potential social and cultural problems the we are going to face. There are no normal humans in the novel that we can see in these days. This paper will highlight how the different transhumans and posthumans interact in social and cultural hi-tech society and what the author thinks the machinic and artificial world will need to be a desirable one. Except for a few characters like Tessier and Marie-France, most of them could be called cyborg, robot, and A.I. Even Case, a cyberspace cowboy, who had a human body is forced to carries some vital chips in his body after he made a fatal mistake. Julius Deane is ?one hundred and thirty-five years old, his metabolism assiduously warped by a weekly fortune in serums and hormones? and ?the code of his DNA? reset by genetic surgeons in Chiba. As a warrior, Molly is a kind of cyborg, whose glasses are ?surgically inset, sealing her sockets? while Riviera is ?a product of the rubble rings that fringe the radioactive core of old Bonn.? Armitage?s memories are artificially operated and implanted into Willis Corto?s body while bodiless Flatline is a computer human downloaded with McCoy Pauley?s brain. Hideo is ?almost certainly the ninja clone? whose ?brown chest was bare and smooth.? As the children of Tessier and Marie-France, 3Jane and 8Jean are probably cloned ones who can inherit forever their parents? immortal global corporation. Wintermute, a cold and goal oriented Artificial Intelligence, is ?hive mind, decision maker, effecting change in the world outside? while Neuromancer is ?personality? and ?immortality.? William Gibson?s Neuromancer reminds us of George Orwell?s work, 1984, depicting the future world that he imagined in the 1940s. William Gibson suggests diverse types of humans who will play a pivotal role in the future world. It is needless to say that cyborgs and robots along with A.I. - the posthumans and transhumans not different from us today ? are expected to actively engage in normal social and cultural life. Nevertheless, it seems that the author would like to emphasize the combination of the warm personality and cold rationality even in the machinic and artificial world which Marie-France would materialize by ?the compulsion that had driven the thing(Wintermute) to free itself, to unite with Neuromancer.? |
| Keywords: | Artificial Intelligence, Cyborg, Robot, Neuromancer, Wintermute, Posthumans and Transhumans |
| Date: | 2018–06 |
| URL: | http://d.repec.org/n?u=RePEc:sek:iacpro:7209004&r=big |