Abstract:
Our research examines a predictive machine learning approach for financial news articles analysis using several different textual representations: bag of words, noun phrases, and named entities. Through this approach, we investigated 9,211 financial news articles and 10,259,042 stock quotes covering the S&P 500 stocks during a five week period. We applied our analysis to estimate a discrete stock price twenty minutes after a news article was released. Using a support vector machine (SVM) derivative specially tailored for discrete numeric prediction and models containing different stock-specific variables, we show that the model containing both article terms and stock price at the time of article release had the best performance in closeness to the actual future stock price (MSE 0.04261), the same direction of price movement as the future price (57.1% directional accuracy) and the highest return using a simulated trading engine (2.06% return). We further investigated the different textual representations and found that a Proper Noun scheme performs better than the de facto standard of Bag of Words in all three metrics.
PMID: 18654475;Abstract:
An analysis of 30 years of data on patent publications from the US Patent and Trademark Office, the European Patent Office and the Japan Patent Office confirms the dominance of companies and selected academic institutions from the US, Europe and Japan in the commercialization of nanotechnology. © 2008 Nature Publishing Group.
Abstract:
As multinational organizations increasingly use the Web to seek information, there is a need for better support of searching the Web across different regions. However, support for Internet searching in non-English speaking regions is much weaker than that in English-speaking regions. To alleviate the problems, we propose a knowledge portal approach to supporting cross-regional searching of multinational organizations. The approach was used to build two Web portals in the Spanish business and Arabic medical domains. Experimental results show that our portals achieved significantly better performance (in terms of search accuracy and user satisfaction) than existing search engines in the corresponding domains. The encouraging findings point to a promising future of the approach to facilitating cross-regional searching in multinational organizations.
Abstract:
Much of the work in business and academia is performed by groups of people. While significant advancement has been achieved in enhancing individual productivity by making use of information technology, little has been done to improve group productivity. Prior research suggests that we should know more about individual differences among group members as they respond to technology if we are to develop useful systems that can support group activities. We report results of a cognitive study in which researchers were observed performing three complex information entry and indexing tasks using an Integrated Collaborative Research System. The observations have revealed a taxonomy of knowledge and cognitive processes involved in the indexing and management of information in a research collaboration environment. A detailed comparison of knowledge elements and cognitive processes exhibited by senior researchers and junior researchers has been made in this article. Based on our empirical findings, we have developed a framework to explain the information management process during research collaboration. Directions for improving design of Integrated Collaborative Research Systems are also suggested. © 1992.
