Lithuanian Data Entry & Quality Control in the Social Media sector

We are looking for someone who speaks English and Lithuanian. You will help us with setting up the Crowdynews Article Widget in Lithuanian. The Crowdynews’ Article Widget matches an article’s content with the most relevant social media content. The widget uses highly advanced language processing to find the main topic of your article and fully automated displays only social content that ishighly relevant and appropriate A. Initial Tasks We need someone to create these things: Stoplist A stoplist is a list of filler words that are added to a language but don't have any real meaning and/or don't contribute anything useful to the text. So for english, typical stopwords would be words like 'this', 'the', 'that', 'why', 'how', 'what', etc. Also days of the weeks (monday-sunday), months of the year (January-February), count words (one, two, three, first, second, third, ...) are words we in general ignore because they don't add anything useful to our keyword analysis. Profanity list We don't want any offensive content to appear in our products. That's why we have a blocklist to filter out profane words, swear words and other undesired words. Stemmer {Maybe} The stemmer is a part of the system that brings back words to its stem ('banker', 'banking' and 'banks' are brought back to 'bank', for example). We don't have any stemming for Afrikaans. It's not strictly essential but can be used to improve relevancy. For this we'd need somebody knowledgeable in the language that can inform us of the proper rules, which we'd then have to implement into a stemmer. We can skip this step initially and work with the stoplist and profanity list and see how that goes. B. Later Task (probably) Doing quality control. That is, are the tweets that are being displayed in the widget relevant? Next week I will be reviewing all the bids. Thank you for your attention. Mathijs ter Wee Social Media Consultant