| # KEYWORD QUERY VS STATEMENT/QUESTION CLASSIFIER FOR NEURAL SEARCH | |
| --- | |
| language: "en" | |
| tags: | |
| - neural-search-query-classification | |
| - neural-search | |
| widget: | |
| - text: "keyword query." | |
| --- | |
| | Train Loss | Validation Acc.| Test Acc.| | |
| | ------------- |:-------------: | -----: | | |
| | 0.000806 | 0.99 | 0.997 | | |
| ```python | |
| from transformers import AutoTokenizer, AutoModelForSequenceClassification | |
| tokenizer = AutoTokenizer.from_pretrained("shahrukhx01/bert-mini-finetune-question-detection") | |
| model = AutoModelForSequenceClassification.from_pretrained("shahrukhx01/bert-mini-finetune-question-detection") | |
| ``` | |
| Trained to add feature for classifying queries between Keyword Query or Question + Statement Query using classification in [Haystack](https://github.com/deepset-ai/haystack/issues/611) | |
| Problem Statement: | |
| One common challenge that we saw in deployments: We need to distinguish between real questions and keyword queries that come in. We only want to route questions to the Reader branch in order to maximize the accuracy of results and minimize computation efforts/costs. | |
| Baseline: | |
| https://www.kaggle.com/shahrukhkhan/question-v-statement-detection | |
| Dataset: | |
| https://www.kaggle.com/stefanondisponibile/quora-question-keyword-pairs | |
| Kaggle Notebook: | |
| https://www.kaggle.com/shahrukhkhan/question-vs-statement-classification-mini-bert/ | |