![\"Photo](\"https://storage.googleapis.com/kaggle-media/competitions/kaggle/3586/media/umbrellas.jpg\")
You cannot sign up to Kaggle from multiple accounts and therefore you cannot submit from multiple accounts.
\r\n\r\n Privately sharing code or data outside of teams is not permitted.\r\n It's okay to share code if made available to all participants on the forums.\r\n
\r\n\r\n Team mergers are allowed and can be performed by the team leader. In order to merge, the combined team must have a total submission count less than or equal to the maximum allowed as of the merge date. The maximum allowed is the number of submissions per day multiplied by the number of days the competition has been running.
\r\n
There is no maximum team size.
\r\nYou may submit a maximum of 5 entries per day.
\r\nYou may select up to 2 final submissions for judging.
\r\nStart Date: 9/27/2013 2:43 PM UTC
\r\nMerger Deadline: None
\r\nEntry Deadline: None
\r\nEnd Date: 12/1/2013 11:59 PM UTC
\r\n\t\r\nThe training set contains tweets, locations, and a confidence score for each of 24 possible labels. The 24 labels come from three categories: sentiment, when, and kind. Human raters can choose only one label from the \"sentiment\" and \"when\" categories, but are allowed multiple choices for the \"kind\". Your goal is to predict a confidence score of all 24 labels for each tweet in the test set.
\r\ns1,\"I can't tell\"\r\ns2,\"Negative\"\r\ns3,\"Neutral / author is just sharing information\"\r\ns4,\"Positive\"\r\ns5,\"Tweet not related to weather condition\"\r\n
\r\nw1,\"current (same day) weather\"\r\nw2,\"future (forecast)\"\r\nw3,\"I can't tell\"\r\nw4,\"past weather\"
\r\nk1,\"clouds\"\r\nk2,\"cold\"\r\nk3,\"dry\"\r\nk4,\"hot\"\r\nk5,\"humid\"\r\nk6,\"hurricane\"\r\nk7,\"I can't tell\"\r\nk8,\"ice\"\r\nk9,\"other\"\r\nk10,\"rain\"\r\nk11,\"snow\"\r\nk12,\"storms\"\r\nk13,\"sun\"\r\nk14,\"tornado\"\r\nk15,\"wind\"\r\n
For example, a tweet \"The hot and humid weather yesterday was awesome!\" could have s4=1, w4=1, k4=1, k5=1, with the rest marked as zero.
\r\nEach tweet is reviewed by multiple raters, and some amount of disagreement on the labels is expected. The confidence score accounts for two factors - the mixture of labels that the raters gave a tweet and the individual trust of each rater. Since some raters are more accurate than others (e.g. they pay closer attention, take the job more seriously, etc.), these raters count more in the confidence score. You have more confidence that a tweet was referring to the past if you trust the person telling you.
\r\nIn this competition you do not have access to the individual ratings or the raters' trust. This \"unknown trust\" issue is therefore a source of noise in the problem. However, you do know that the raters can choose only one label from the \"sentiment\" and \"when\" categories, but multiple choices for the \"kind\". The result is that confidences for the \"sentiment\" and \"when\" categories sum to one. Conversely, the sum of the \"kind\" category will not always sum to one.
", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Description", "content": "
In this competition you are provided a set of tweets related to the weather. The challenge is to analyze the tweet and determine whether it has a positive, negative, or neutral sentiment, whether the weather occurred in the past, present, or future, and what sort of weather the tweet references. It's a lot to mine from so few characters, but if the going gets tough you can always blame the weather...
\r\n\"Please knock out the power giant storm that is passing thru....please.\" -Tweet #74096
\r\nWe are excited to team up with CrowdFlower on the first of what we hope will be many fun machine learning projects. CrowdFlower is debuting a new open data library and we're always looking for an excuse to have a competition. Why is this exciting? Sweet, sweet Labels.
\r\nData repositories sometimes have more in common with a landfill than a library. They're home to tattered piles of spreadsheets in odd formats with nary a shred of documentation to tell the GDP of Chile from the migratory patterns of North American goldfinches. If creating value from this digital exhaust is a defining theme of the big data explosion, most repositories leave you choking on the diesel fumes of data disappointment. Such data is great if you are doing a report on the GDP of Chile, but not so useful if you are doing machine learning, or its red-headed step child, data science.
\r\nCrowdflower's data sets provide the thing that makes so many repositories fall short - data paired with labels. One can decide whether two English sentences are related, make judgments about yogurt chatter, or rank emotions on tweets about nuclear energy. It's all about the (wo)manpower to label what these bytes actually mean.
\r\nCrowdFlower Open Data Library is a repository of real data set samples that developers, researchers and data scientists can download and use to test and improve algorithms. Our mission is to encourage users to explore the possibilities and power of crowdsourcing. Open Data is free, available to anyone, and ready-to-use with CrowdFlower’s Platform.
\r\nNew data sets are continuously added to CrowdFlower Open Data Library as users of the CrowdFlower Platform opt-in to share their data with the crowdsourcing community. Sample data sets currently available include tweets for sentiment and topic analysis, word combinations to test similarities, sentence combinations to test related topics, and more. Learn more at www.crowdflower.com.
", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Prizes", "content": "The winning team will be rewarded $500 of credit on the Crowdflower platform. It is strongly encouraged (bordering on peer pressure) that this is used to create a dataset for the next Crowdflower research competition on Kaggle. We hope that this initiative can generate great problems and new datasets for everyone to enjoy.
\r\nIf the winning team is not interested in creating/labeling a dataset, we will continue to offer the prize to the highest team that is interested.
", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Evaluation", "content": "The Root Mean Squared Error (\"RMSE\") is used to measure the accuracy:
\r\n\\[ \\textrm{RMSE} = \\sqrt{\\frac{\\sum_{i=1}^n (p_i - a_i)^2}{n} } \\]
\r\nWhere:
\r\nFor each tweet in the test set, predict the confidence score for each of the 24 possible labels. Submission files should be in the following format and must have a header.
\r\nid,s1,s2,s3,s4,...,k12,k13,k14,k15\r\n4,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0\r\n5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0\r\n7,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0\r\netc.\r\n", "isPublished": false, "hasIsPublished": false, "postTitle": "", "hasPostTitle": false, "mimeType": "", "hasMimeType": false }, { "name": "Winners", "content": "
First Place:
\r\n