BEER: BEtter Evaluation as Ranking
نویسندگان
چکیده
We present the UvA-ILLC submission of the BEER metric to WMT 14 metrics task. BEER is a sentence level metric that can incorporate a large number of features combined in a linear model. Novel contributions are (1) efficient tuning of a large number of features for maximizing correlation with human system ranking, and (2) novel features that give smoother sentence level scores.
منابع مشابه
The Presentation of an Approach of Evaluation and Ranking in Data Envelopment Analysis with Interval Data: a Case Study in the Evaluation and Ranking of Iran’s Provinces in the Health and Treatment Sector
Today, in every society the health and treatment sector are among the most important service sectors. Therefore, it is crucial that their performance be evaluated and examined. Although the researchers have proposed many different approaches to evaluate and rank the health sectors, no precise approach for evaluating and ranking have been reported up to now. Assessing the coefficient of variatio...
متن کاملTemporal Factors in Beer Ratings CSE 255 , Winter 2014 Project 1
Predicting a user’s rating of a product without taking how their preferences might change over time leaves out a lot of data that could be used to improve that prediction. Why not track that change in preferences and change the recommendations accordingly? As long as keeping track of that doesn’t add significantly to the model there should be no reason not to. In this paper, I aim to do precise...
متن کاملEvaluation and ranking of suppliers with fuzzy DEA and PROMETHEE approach
Supplier selection is a multi-Criteria problem. This study proposes a hybrid model for supporting the suppliers’ selection and ranking. This research is a two-stage model designed to fully rank the suppliers where each supplier has multiple Inputs and Outputs. First, the supplier evaluation problem is formulated by Data Envelopment Analysis (DEA), since the regarded decision deals with uncertai...
متن کاملPreconcentration and Determination of Trace Amounts of Aromatic Compounds in Non-alcoholic Beer Samples by In-syringe Dispersive liquid–liquid Microextraction
In this paper, a new simple and effective method based on dispersive liquid–liquid microextraction procedure is proposed for rapid and simultaneous separation and preconcentration of ultra trace amounts of benzene, toluene, ethyl benzene and xylenes (BTEX) in non-alcoholic Beer samples. In this experiment, a glass syringe was used as extraction unit. With this simple configuration, the centrifu...
متن کاملRanking of units by anti-ideal DMU with common weights
Data envelopment analysis (DEA) is a powerful technique for performance evaluation of decision making units (DMUs). One of the main objectives that is followed in performance evaluation is discriminating among efficient DMUs to provide a complete ranking of DMUs. DEA successfully divides them into two categories: efficient DMUs and inefficient DMUs. The DMUs in the efficient category have ident...
متن کامل