Intelligent Document Querying with LLMs and Speech Input: A Scalable Semantic Framework
Bibliographic record
Abstract
The information era has created challenges in terms of retrieving domain-specific information from voluminous unstructured documents (like Education and Businesses), whether via informer attacks or database attacks. Researchers have developed an information framework that includes LlamaIndex - which does intelligent document analysis; Gradient AI - which provides superior inferences; and Astra DB - which provides the appropriate storage solutions for these systems. More recently, we have expanded our research to include varied modalities of multi-modal data such as text and voice, using different configurations, e.g., BGE for embeddings, Wav2Vec2 for speech. We verified that the use of LlamaIndex, Gradient AI and Astra DB on varying document sets produced results averaging 2.7 seconds per query with an average of the “top 3” results indicating a semantic search performance rate of 91.2%. Overall, the proposed framework, when compared to existing technologies such as LangChain (88.4%) and Haystack$(86.9 \%)$, has produced the highest level of performance. The response to user queries has been done in an informative and concise manner inline with the context and has been achieved via the use of RESTful APIs. The proposed model is ready to be tested in real-world applications and will be made available for additional trials beginning early next year.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.004 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".