Sentiment diversification with different biases

Abstract

Prior search result diversification work focuses on achieving topical variety in a ranked list, typically equally across all aspects. In this paper, we diversify with sentiments according to an explicit bias. We want to allow users to switch the result perspective to better grasp the polarity of opinionated content, such as during a literature review. For this, we first infer the prior sentiment bias inherent in a controversial topic -the 'Topic Sentiment'. Then, we utilize this information in 3 different ways to diversify results according to various sentiment biases: (1) Equal diversification to achieve a balanced and unbiased representation of all sentiments on the topic; (2) Diversification towards the Topic Sentiment, in which the actual sentiment bias in the topic is mirrored to emphasize the general perception of the topic; (3) Diversification against the Topic Sentiment, in which documents about the 'minority' or outlying sentiment(s) are boosted and those with the popular sentiment are demoted.Since sentiment classification is an essential tool for this task, we experiment by gradually degrading the accuracy of a perfect classifier down to 40%, and show which diversification approaches prove most stable in this setting. The results reveal that the proportionality-based methods and our SCSF model, considering sentiment strength and frequency in the diversified list, yield the highest gains. Further, in case the Topic Sentiment cannot be reliably estimated, we show how performance is affected by equal diversification when actually an emphasis either towards or against the Topic Sentiment is desired: in the former case, an average of 6.48% is lost across all evaluation measures, whereas in the latter case this is 16.23%, confirming that bias-specific sentiment diversification is crucial.

Tag-Based Paper Retrieval: Minimizing User Effort with Diversity Awareness

“…It aims to improve user satisfaction by providing a diverse view of information, thereby increasing the probability of returning some information that truly matches the user's expectation. Various applications that have benefited from diversification include sentiment analysis [4], web search [12], database search [7], largescale visualization [31], social network [40] and recommender systems [11]. In our case, since users cannot often precisely and exhaustively describe their queries, increasing diversity of tag-query suggestion will provide users more chances to find the desired papers quickly.…”

Learning Ranked Sentiment Lexicons

“…Martineau et al [14] proposed the metric D-TFIDF to weight words scores. Martineau et al found that D-TFIDF improved the classification accuracy of subjective sentences in the Pang and Lee subjectivity dataset 1 . Moreover, variants of our proposed method outperformed the static lexicons with D-TFIDF showing similar performance.…”

“…Later, Jo and Oh [13] proposed a unified aspect and sentiment model based on the assumption that each sentence concerns one aspect and all sentiment words in that sentence refer to that sentence. Finally, Aktolga and Allan [1] targeted the task of sentiment diversification in search results. The common element among these works [1,9,13,28] is the use of the sentiment lexicon SentiWordNet [3].…”

“…In this new context, sentiment expressions or opinion expressions, are important pieces of information, specially, in the context of online commerce [12]. Therefore, modeling text to find meaningful words for expressing sentiments (sentiment lexicons) emerged as an important research direction [1,9,13,28]. In this work we investigate the viability of automatically generating a sentiment lexicon for opinion retrieval and sentiment classification applications.…”

“…Finally, Aktolga and Allan [1] targeted the task of sentiment diversification in search results. The common element among these works [1,9,13,28] is the use of the …”