Working with Crowdsourced Textual Data