Definitions Properties

The list of properties described here is contained in file: definitions.properties. They control internal functionalities of the benchmark e.g. the distributions of queries or entities in generated data. It is not recommended to modify that file.

• aboutsAllocations - defines allocation amount of About tags in Creative Works
• mentionsAllocations - defines allocation amount of Mention tags in Creative Works
• entityPopularity - defines popularity of an entity in the reference datasets
• usePopularEntities - defines allocation amount of popular entities to be used when tagging in Creative Works or in aggregation queries. Used for generation of Creative Works biased towards popular entities
• creativeWorkTypesAllocation - defines allocation amount of Creative Work Types : BlogPost, NewsItem, Programme
• aboutAndMentionsAllocation - defines allocation amount of about or mentions used for the main aggregation query (/data/sparql/aggregation/query1.txt), which one will be used more frequently
• editorialOperationsAllocation - defines allocation amount of queries in the editorial query mix that each editorial agent will execute. Query mix order : insert.txt, update.txt and delete.txt
• aggregationOperationsAllocation - defines allocation amount of queries in the aggregation query mix that each aggregation agent will execute. Query mix order : query1.txt, query2.txt... etc
• exponentialDecayUpperLimitOfCWs - defines the maximum number of creative works that an entity can be tagged about. Exponential decay function will start from the value defined
• exponentialDecayRate - defines the exponential decay rate. Used values to be in range 0.01 (for gentle slope) to 1 (for steep slope)
• exponentialDecayThresholdPercent - defines the threshold in percents of exponential decay, below that threshold values will be ignored. Threshold is defined as the ratio of : currentExponentialDecayValue / exponentialDecayUpperLimitOfCWs. e.g. 5% threshold will be the value of 0.05
• majorEvents - defines the maximum number of 'major' events that could happen during data generation period. Each major event will be tagged by a number of Creative Works which will decay exponentially in time
• minorEvents - defines the maximum number of 'minor' events that could happen during data generation period. Each minor event will be tagged by a number of Creative Works which will decay exponentially in time. Value of exponentialDecayUpperLimitOfCWs for minor events will be ten times smaller for them
• seedYear - defines a seed year that will be used for generating the Creative Works. Each Creative Work will have its creation date during that year. All date-range queries will use that value also
• dataGenerationPeriodYears - defines the period (in years) of the gnerated data, starting from 'seedYear'
• correlationsAmount - defines the amount of correlations that data generator will model between entities from reference knowledge data. Default value (50) will produce around 15 million triples
• correlationsMagnitude - defines maximum amount of Creative Works that will be generated for a particular correlation in a single day
• correlationDuration - defines the duration of correlation between two entities as a percent of the total data generation period. Default generation period is one year
• correlationEntityLifespan - defines the life span of each entity that participates in a correlation as a percent of the total data generation period
• minLat - efines minimum latitude, a geo-spatial property used to configure the geo-spatial search area of queries
• maxLat - defines maximum latitude, a geo-spatial property
• minLong - defines minimum longtitude, ,a geo-spatial property
• maxLong - defines maximum longtitude, a geo-spatial property
• mileStoneQueryPosition -
• maxLat - defines the position in terms of percents at which a milestone query is executed (related to Online and Replication Benchmark feature)
• queryPools - defines pools of queries, where each pool contains a unique set of queries. During query execution, each query from a pool gets executed just once until all queries in the pool have been executed. Each query pool is defined by a set of curly braces {}. If empty value has been assigned to the queryPools property, then no query pools are created, all queries are executed according to distributions defined in parameter 'aggregationOperationsAllocation'