Hybrid approaches in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to combine rule-based and machine learning predictions.

NLP

final_prediction = ml_model.predict(data) if [1] else rule_based_predict(data)

Drag options to blanks, or click blank then click option'

Ause_ml

Buse_rule

Cuse_hybrid

Duse_data

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'use_rule' instead of 'use_ml' causes the rule-based method to always run.

Using 'use_data' is incorrect because it does not control the prediction method.

The variable 'use_ml' controls whether to use the machine learning model's prediction or the rule-based method.

2fill in blank

medium

Complete the code to merge predictions from ML and rules by averaging.

NLP

combined_score = (ml_score + [1]) / 2

Drag options to blanks, or click blank then click option'

Aml_prediction

Brule_prediction

Cdata_score

Drule_score

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'ml_prediction' duplicates the ML score instead of combining with rules.

Using 'data_score' is unrelated to rule-based predictions.

We average the machine learning score and the rule-based score to get a combined score.

3fill in blank

hard

Fix the error in the hybrid prediction function by filling the blank.

NLP

def hybrid_predict(data):
    if data['length'] > 10:
        return ml_model.predict(data)
    else:
        return [1](data)

Drag options to blanks, or click blank then click option'

Adata_predict

Bml_model.predict

Crule_based_predict

Dpredict_rule

Attempts:

3 left

💡 Hint

Common Mistakes

Calling 'ml_model.predict' again ignores the rule-based fallback.

Using 'predict_rule' is not defined and causes an error.

The else branch should call the rule-based prediction function named 'rule_based_predict'.

4fill in blank

hard

Fill both blanks to create a dictionary of predictions from ML and rules.

NLP

predictions = {
    'ml': ml_model.[1](data),
    'rule': [2](data)
}

Drag options to blanks, or click blank then click option'

Apredict

Brule_based_predict

Cpredict_rule

Dpredict_ml

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'predict_rule' causes a NameError.

Using 'predict_ml' is not a valid method.

The ML model uses 'predict' method, and the rule-based function is 'rule_based_predict'.

5fill in blank

hard

Fill all three blanks to filter data, predict with ML, and combine with rules.

NLP

filtered_data = [d for d in dataset if d['score'] [1] 0.5]
ml_preds = [ml_model.[2](d) for d in filtered_data]
final_preds = [ml + [3](d) for ml, d in zip(ml_preds, filtered_data)]

Drag options to blanks, or click blank then click option'

A>

Bpredict

Crule_based_predict

D<

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' filters wrong data subset.

Using 'predict_rule' instead of 'rule_based_predict' causes errors.

We filter data with '>', predict with ML using 'predict', and add rule-based predictions using 'rule_based_predict'.