The characteristics of agricultural big data satisfy the five characteristics of big data, one is the large volume of data (Volume), the second is the fast processing speed (Velocity), the third is the many types of data (Variety), the fourth is the large value (Value), and the fifth is the high accuracy (Veracity). Including the following:

(1) From the field point of view, the agricultural field as the core (covering planting, forestry, animal husbandry and other sub-industries), gradually expanding to the relevant upstream and downstream industries (feed production, fertilizer production, production of agricultural machinery, slaughtering, meat processing industry, etc.), and integrating the macro-economic background of the data, including statistical data, import and export data, price data, production data and even meteorological data. The data will be integrated into the macroeconomic context, including statistics, import and export data, price data, production data, and even weather data.

(2) From the geographical point of view, taking domestic regional data as the core, and drawing on international agricultural data as an effective reference; not only including national-level data, but also covering provincial and municipal data, and even local and municipal data, to provide the basis for accurate regional research;

(3) From the point of view of granularity, not only statistical data, but also basic information on agricultural economic entities, investment information, shareholder information, patent information, import and export information, and even meteorological data. information, patent information, import and export information, recruitment information, media information, GIS coordinate information and so on.

(4) From the point of view of specialization, it should be implemented step by step, first of all, it is to build the professional data resources in the field of agriculture, and secondly, it should gradually and orderly plan the professional data resources in the sub-fields, such as the professional monitoring data for the livestock breeds of hogs, broilers, egg-laying chickens, beef cows, dairy cows, and meat goats, and so on.