Search results

The purpose of this paper is to describe evaluation data and a prototype system named WISDOM used for analyzing information credibility based on natural language processing.

The authors started the Information Credibility Criteria project in April, 2007, mainly to analyze the credibility of information (text) on the web. The project proposes to capture information credibility based on four criteria (content, sender, appearance, and social valuation) and aims to analyze and organize them logically using natural language processing based on predicate argument structure.

The evaluation data described in this paper were developed as learning and verifying data for these various analysis modules and are composed of manually‐annotated data based on each evaluation criteria about several pre‐selected topics such as current events and medical issues. The prototype system WISDOM was developed to provide information credibility from different perspectives.

Users will be able to find credible information more reliably by browsing information using different evaluation criteria and conditions provided by the system.

To develop a method for classifying information sender of web documents, which constitutes an important part of information credibility analysis.

Machine learning approach was employed. About 2,000 human‐annotated web documents were prepared for training and evaluation. The classification model was based on support vector machine, and the features used for the classification included the title and URL of documents, as well as information of the top page.

With relatively small set of features, the proposed method achieved over 50 per cent accuracy.

Some of the information sender categories were found to be more difficult to classify. This is due to the subjective nature of the categories, and further refinement of the categories is needed.

When combined with opinion/sentiment analysis techniques, information sender classification allows more profound analysis based on interactions between opinions and senders. Such analysis forms a basis of information credibility analysis.

This study formulated the problem of information sender classification. It proposed a method which achieves moderate performance. It also identified some of the issues related to information sender classification.

This paper aims to report a QA system that can answer how‐type questions based on confirmed knowledge acquired from mails, posted to a mailing list. It aims to propose a method of detecting incorrect information in mails posted to a mailing list (ML) by using mails that ML participants submitted for correcting incorrect information in previous mails.

The paper discusses a problem of acquiring knowledge from natural language documents, then proposes a method to give these mails three kinds of confirmation labels, positive, negative, and other, depending on their credibility.

The paper shows a QA system based on the confirmed knowledge. It finds mail questions that are similar to the user's question and gives answers and their confirmation labels to the user. By using the confirmation labels, the user can easily choose the information that can solve his or her problem.

The study describes a method of detecting incorrect information in mails posted to a mailing list and acquiring confirmed knowledge from them.

The purpose of this paper is to propose a new direction for managing information‐credibility risk in the current information and communications technology (ICT) era, where ICT has had both positive and negative effects on contemporary society.

The paper takes a practical and inductive approach to study the Kyoto avian influenza panic and countermeasures taken in 2004.

The paper identifies factors which led to enormous damage through harmful rumors and proposes new perspectives for devising countermeasures, such as increasing consumer confidence in an agency as a source of information and effective management of knowledge transfer from experts to non‐experts.

The study gains a better understanding of both technological and social factors that enable or detract from effective nationwide management of information‐credibility risk. Many related ICT projects have been based on either human resource systems or advanced technology. It considers the integration of both factors from three perspectives.

This is a new perspective for examining the transfer of knowledge from experts to consumers in terms of practical solutions, in contrast to the many existing knowledge‐related articles that have mainly focused on knowledge management among experts.

To evaluate and extend, existing natural language processing techniques into the domain of informal online political discussions.

A database of postings from a US political discussion site was collected, along with self‐reported political orientation data for the users. A variety of sentiment analysis, text classification, and social network analysis methods were applied to the postings and evaluated against the users' self‐descriptions.

Purely text‐based methods performed poorly, but could be improved using techniques which took into account the users' position in the online community.

The techniques we applied here are fairly simple, and more sophisticated learning algorithms may yield better results for text‐based classification.

This work suggests that social network analysis is an important tool for performing natural language processing tasks with informal web texts.

This research extends sentiment analysis to a new subject domain (US politics) and a new text genre (informal online discusssions).