The measurement of meaning

Giving Content to Investor Sentiment: The Role of Media in the Stock Market

We investigate whether investor sentiment affects the cross-section of stock returns in China A-share market.The evidence shows that the sensitivity of stock returns to sentiment changes is different. When an index of investor sentiment takes high values,low tangible assets,high Debt-asset ratio,and non-dividend-paying stocks earn relatively higher returns,when sentiment is low,the aforementioned categories of stocks earn relatively lower returns.When sentiment is high,low-price,unprofitable and high book-market ratio stocks earn relatively higher returns and vice versa,but which is insignificant.The capitalization, volatility and institutional ownership appear to have no significant cross-section effect of investor sentiment on its characteristic Portfolio return.

Investor Sentiment and the Cross-Section of Stock Returns

During natural and man-made disasters, people use social media platforms such as Twitter to post textual and multimedia content to report updates about injured or dead people, infrastructure damage, missing or found people, among other information types. Studies have revealed that this online information, if processed timely and effectively, is extremely useful for humanitarian organizations to gain situational awareness and plan relief operations. In addition to the analysis of textual content, recent studies have shown that imagery content on social media can boost disaster response significantly. Despite extensive research that mainly focuses on textual content to extract useful information, limited work has focused on the use of imagery content or the combination of both content types. One of the reasons is the lack of labeled imagery data in this domain. Therefore, in this paper, we aim to tackle this limitation by releasing a large multi-modal dataset from natural disasters collected from Twitter. We provide three types of annotations, which are useful to address a number of crisis response and management tasks for different humanitarian organizations.

/pdf/crisismmd-multimodal-twitter-datasets-from-natural-disasters-4hj9o5haow.pdf

CrisisMMD: Multimodal Twitter Datasets from Natural Disasters

Twitter as a tool for the management and analysis of emergency situations: A systematic literature review

The influence of news on financial markets has been studied by extracting opinions and sentiment from text using content analysis and natural language processing techniques, and then using this measure to estimate the impact of sentiment on price changes. We present a method and implementation that analyses the content of news using multiple dictionaries that accounts for the specific use of terminology in a given domain. To evaluate our approach we build different collections of domain related news for two financial markets and examine the impact that topical news has on two financial benchmarks in the equity and oil markets. We examine how the level of news sentiment from different news sources influences financial returns over time. We create a trading signal based on the news impact that predicts next day returns in the Dow Jones Industrial Average and West Texas Intermediate crude oil. We find that incorporating news sentiment into a trading strategy increases annual returns over a simple buy and hold strategy for both markets.

Estimating the impact of domain-specific news sentiment on financial assets

Distributed morality, privacy, and social media in natural disaster response

Social media platforms have become a source of high volume, real-time information describing significant events in a timely fashion. In this paper we describe a system for the real-time extraction of information from text and image content in Twitter messages and combine the spatio-temporal metadata of the messages to filter the data stream for emergency events and visualize the output on an interactive map. Twitter messages for a geographic region are monitored for flooding events by analysing the text content and images posted. Events detected are compared with a ground truth to see if information in social media correlates with actual events. We propose an Intrusion Index as part of this prototype to facilitate ethical harvesting of data. A map layer is created by the prototype system that visualises the analysis and filtered Twitter messages by geolocation.

Mining Multimodal Information on Social Media for Increased Situational Awareness.

Bonferroni mean aggregates the interaction between all pairs of inputs from some $n$ -dimensional input vector. Therefore, it is able to capture the dependency structure between the inputs. Weighted version of the Bonferroni mean then assumes that each input has a possibly different weight. Such an approach poses several constraints that decrease the modeling flexibility of the dependency structure on the input space. In this paper, we present an overview of the different approaches to weighted Bonferroni mean and introduce the Bonferroni mean with weighted interaction where interactions between inputs are weighted rather than the inputs themselves. We show several important properties of the Bonferroni mean with weighted interaction and its connection to concepts studied before. Finally, using the described approach we design a system for emergency management that is able to raise an alarm in the case where certain criteria are met.

Bonferroni Mean With Weighted Interaction

Flood event monitoring plays an important role for emergency management. With the fast growth of social media, a large number of images and videos are uploaded and searched on the internet during disasters, which can be used as “sensors” for improving efficiency of emergency management. This work proposes a novel framework in which the rich information available from social media is incorporated with image analysis to enhance image retrieval for disaster management. The text associated with images of flooding events was used to extract prominent words associated with flooding. The image features are represented by a histogram of visual words obtained using the Bag-of-Words (BoW) model. The text and image analysis are integrated at the feature level, in which the text features are conjoined directly with image features. The proposed approach was evaluated based on two flood event corpuses obtained from the US Federal Emergency Management Agency media library and public Facebook pages and groups related to flood and flood aid (in German). The experimental results demonstrate the improved performance of image recognition after incorporating the text features, which suggests the potential to enhance the efficiency of emergency management.

/pdf/integration-of-text-and-image-analysis-for-flood-event-image-1vl1vu81zp.pdf

Stephen Kelly

Papers

Estimating the impact of domain-specific news sentiment on financial assets

Distributed morality, privacy, and social media in natural disaster response

Mining Multimodal Information on Social Media for Increased Situational Awareness.

Bonferroni Mean With Weighted Interaction

Integration of text and image analysis for flood event image recognition