As illustrated in FIG. 4, model 400 can comprise an input layer 402, an embedding layer 404, and/or an output layer 406. In the example trading application embodiment described above with reference to FIG. 1, input layer 402 can provide to embedding layer 404 a Java class (denoted as “MarketSummarySingleton” in FIG. 4) of a monolithic application in the form of an input vector as depicted in FIG. 4. Embedding layer 404 can produce one or more embeddings of the input Java class, where such one or more embeddings can comprise one or more vector representations of the input Java class. Based on such vector representation(s) of the input Java class, output layer 406, which can comprise a softmax classifier as illustrated in FIG. 4, can predict whether another class of the monolithic application is in a predefined window (e.g., a window size of 3 as denoted in the example trading application embodiment described above with reference to FIG. 1). For example, output layer 406 can predict whether a pair of Java classes in a monolithic application belong to the same window (denoted as “the neighborhood” in FIG. 4). For instance, in the example trading application embodiment described above with reference to FIG. 1 and as illustrated in FIG. 4, based on embedding layer 404 producing the vector representations of the input Java class denoted in FIG. 4 as “MarketSummarySingleton,” output layer 406 can predict: a) the probability that the class “Log” is in the neighborhood (e.g., within a window size of 3 with respect to the input Java class “MarketSummarySingleton”); b) the probability that the class “TradeConfig” is in the neighborhood (e.g., within a window size of 3 with respect to the input Java class “MarketSummarySingleton”); c) the probability that the class “QuoteDataBeam” is in the neighborhood (e.g., within a window size of 3 with respect to the input Java class “MarketSummarySingleton”); and/or d) the probability that the class “TradeSLSBBeam” is in the neighborhood (e.g., within a window size of 3 with respect to the input Java class “MarketSummarySingleton”).
