https://github.com/stefanoghinelli/salton
Information Retrieval class project, an IR system built upon a corpus of research papers. It ranks results using the BM25 function
https://github.com/stefanoghinelli/salton
bm25 information-retrieval nltk okapi python unimore-informatica whoosh
Last synced: about 1 hour ago
JSON representation
Information Retrieval class project, an IR system built upon a corpus of research papers. It ranks results using the BM25 function
- Host: GitHub
- URL: https://github.com/stefanoghinelli/salton
- Owner: stefanoghinelli
- Created: 2023-07-20T20:11:35.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-09-30T17:00:17.000Z (about 1 year ago)
- Last Synced: 2024-12-29T23:45:14.897Z (10 months ago)
- Topics: bm25, information-retrieval, nltk, okapi, python, unimore-informatica, whoosh
- Language: Python
- Homepage:
- Size: 172 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# [salton](https://en.wikipedia.org/wiki/Gerard_Salton)
🚧 
🚧
## Project description
Salton is a vertical search engine built on a corpus of documents sourced from [CORE](https://core.ac.uk), a public repository of open-access research papers.
The goal is to provide a more refined search experience than the CORE portal.
It uses the [Okapi BM25](https://en.wikipedia.org/wiki/Okapi_BM25) ranking function to estimate the relevance of documents.
End users can formulate queries based on a defined language, results are ranked by relevance with title, score, URL and summary of the abstract.
## Architecture
## Running the project
This project runs using Python 3 and pip. To install it as a Python package, do the following:
1. Clone the repository and change directory
```bash
$ git clone https://github.com/stefanoghinelli/salton.git
$ cd salton
```
2. Install from source
```bash
$ python3 -m venv venv
$ source venv/bin/activate
$ pip install -e .
```
3. Download NLTK corpora
```bash
$ python3 -c "import ssl, nltk; ssl._create_default_https_context=ssl._create_unverified_context; [nltk.download(p) for p in ['punkt','stopwords','omw-1.4','wordnet','averaged_perceptron_tagger']]"
```
4. Setup environment
```bash
$ sh setup_scripts/01.prepare_environment.sh
```
## Command details
```bash
Usage: salton [OPTIONS] COMMAND [ARGS]...
Salton: your tool for retrieving papers faster
Options:
--help Show this and exit.
Commands:
fetch Fetch papers from CORE repository
preprocess Tokenize, lemmatize, remove stopwords
index Build the index
search Query papers by keyword
stats Show statistics
benchmark Run benchmarks
```
### Usage
The project builds Salton locally for command line running.
To fetch papers (100 by default):
```bash
$ salton fetch -l [number of papers]
```
To preprocess papers:
```bash
$ salton preprocess [--wsd]
```
`--wsd`: enables word sense disambiguation (off by default)
> [!NOTE]
> The word sense disambiguation computes similarity between word senses and compares each term against multiple context. This quadratic operation can be highly time consuming.
To build the index:
```bash
$ salton index
```
To query papers by keyword:
```bash
$ salton search -q "[your query]" -l [number of results]
```
To view the statistics:
```bash
$ salton stats
Index statistics:
-Documents indexed: 55
-Unique terms: 17041
-Index size: 8.41 MB
Data statistics:
-Raw papers: 61
-Processed papers: 55
Benchmark statistics:
-Available query sets: 3
```
## Evaluation
### Setup benchmarks
To run benchmarks, you'll need aset of test queries in the `evaluation` directory:
- `query_natural_lang.txt`: natural language queries
- `query_benchmark.txt`: structured queries
- `query_relevance.txt`: relevance data
### Benchmark metrics
The currently supported metrics are precision, recall, NDCG, MAP.
To run benchmarks:
```bash
$ salton benchmark [--save/--no-save] [--detailed/--simple]
```
`--save/--no-save`: results to file (default: save)
`--detailed/--simple`: level of detail in results (default: simple)
## Results
```bash
salton search -q "artificial intelligence" -l 3
==================================================
Results for: artificial intelligence
==================================================
1. Title: SIR A New Wireless Sensor Network Routing Protocol Based on Artificial Intelligence
Score: 23.8540
Abstract: Currently, Wireless Sensor Networks (WSNs) are formed by hundreds of
low energy and low cost micro-electro-mechanical systems. However,
conventional Quality of Service routing models, are not suitable for
ad hoc sensor networks, due to the dynamic nature of such systems.
URL: https://core.ac.uk/download/161255615.pdf
2. Title: A motion system for social and animated robots
Score: 9.9541
Abstract: The social robot Probo is used to study Human-Robot Interactions
(HRI), with a special focus on Robot Assisted Therapy (RAT). The
motion system has a Combination Engine, which combines motion commands
that are triggered by a human operator with motions that originate
from different units of the cognitive control architecture of the
robot.
URL: https://core.ac.uk/download/55844762.pdf
3. Title: On the Collaboration of an Automatic Path-Planner and a Human User for Path-Finding in Virtual Industrial Scenes
Score: 6.3338
Abstract: This paper describes a global interactive framework enabling an
automatic path-planner and a user to collaborate for finding a path in
cluttered virtual environments. The user can then influence the
planner by not following the path and automatically order a new path
research.
URL: https://core.ac.uk/download/12043111.pdf
```