Do personalisation and emotions affect the use of cancer-related websites?

The role of emotional aspects in the information retrieval from the web

Purpose

The purpose of this paper is to identify the role of emotional aspects in information retrieval of PhD students from the web.

Design/methodology/approach

From the methodological perspective, the present study is experimental and the type of study is practical. The study population is PhD students of various fields of science. The study sample consists of 50 students as selected by the stratified purposive sampling method. The information aggregation is performed by observing the records of user’s facial expressions, log file by Morae software, as well as pre-search and post-search questionnaire. The data analysis is performed by canonical correlation analysis.

Findings

The findings showed that there was a significant relationship between emotional expressions and searchers’ individual characteristics. Searchers satisfaction of results, frequency internet search, experience of search, interest in the search task and familiarity with similar searches were correlated with the increased happy emotion. The examination of user’s emotions during searching performance showed that users with happiness emotion dedicated much time in searching and viewing of search solutions. More internet addresses with more queries were used by happy participants; on the other hand, users with anger and disgust emotions had the lowest attempt in search performance to complete search process.

Practical implications

The results imply that the information retrieval systems in the web should identify emotional expressions in a set of perceiving signs in human interaction with computer, similarity, face emotional states, searching and information retrieval from the web.

Originality/value

The results explicit in the automatic identification of users’ emotional expressions can enter new dimensions into their moderator and information retrieval systems on the web and can pave the way of design of emotional information retrieval systems for the successful retrieval of users of the network.

Gender bias in machine learning for sentiment analysis

Purpose

The purpose of this paper is to investigate whether machine learning induces gender biases in the sense of results that are more accurate for male authors or for female authors. It also investigates whether training separate male and female variants could improve the accuracy of machine learning for sentiment analysis.

Design/methodology/approach

This paper uses ratings-balanced sets of reviews of restaurants and hotels (3 sets) to train algorithms with and without gender selection.

Findings

Accuracy is higher on female-authored reviews than on male-authored reviews for all data sets, so applications of sentiment analysis using mixed gender data sets will over represent the opinions of women. Training on same gender data improves performance less than having additional data from both genders.

Practical implications

End users of sentiment analysis should be aware that its small gender biases can affect the conclusions drawn from it and apply correction factors when necessary. Users of systems that incorporate sentiment analysis should be aware that performance will vary by author gender. Developers do not need to create gender-specific algorithms unless they have more training data than their system can cope with.

Originality/value

This is the first demonstration of gender bias in machine learning sentiment analysis.