A team of researchers from the Queensland University of Technology have developed an algorithm to weed out harmful and abusive posts against women on Twitter, says an IANS report. The statistical model was developed by refining a dataset of 1 million tweets by searching for tweets with abusive words like whore, slut and rape. The team's model reportedly identified misogynistic content with 75% accuracy, outperforming previous such models that aimed to identify social interaction.