Targeting recommendation is a hybrid recommendation engine that will recommend targeting attributes to advertisers when they are setting up advertisement campaigns. There are different types of recommendation engines, I choose to build a hybrid recommendation engine because along with user input we will use the historical booking data to recommend targeting attributes. The following steps will be performed to complete this project. 1). As part of this project, I will build ETL (Extraction, Transformation, and Load) pipelines to load new targeting attributes and metadata from smart TV ACR (Automatic Content Recognition) events feed into the new data mart. The new data mart will be used for various analytics purposes and the new targeting attributes and metadata will be used along with the existing targeting attributes for targeting. The ETL pipelines will be developed using big data technologies like Spark for distributed processing, AWS for cloud storage, and Vertica columnar database to build the new data mart based on the data warehousing concepts like hybrid schemas (a combination of Star and Snowflake schemas). 2). Historical bookings are stored in the Postgres transactional DB, the targeting attributes will be extracted and transformed into documents as needed to build the recommendation engine. I will use the Python programming language and Pandas data frames to transform the data. 3). Develop two versions of hybrid recommendation engines in Python using memory-based neighborhood algorithms like cosine similarity and CountVectorizer, and Universal sentence encoder DNN model.