Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/artpar/languagecrunch
LanguageCrunch NLP server docker image
https://github.com/artpar/languagecrunch
coreference-resolution natural-language-processing nlp relation-extraction sentiment-analysis spacy-nlp word2vec wordnet
Last synced: 15 days ago
JSON representation
LanguageCrunch NLP server docker image
- Host: GitHub
- URL: https://github.com/artpar/languagecrunch
- Owner: artpar
- Created: 2018-01-14T17:32:55.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2022-12-07T23:56:51.000Z (almost 2 years ago)
- Last Synced: 2024-10-12T13:17:37.665Z (29 days ago)
- Topics: coreference-resolution, natural-language-processing, nlp, relation-extraction, sentiment-analysis, spacy-nlp, word2vec, wordnet
- Language: Python
- Homepage:
- Size: 589 KB
- Stars: 287
- Watchers: 7
- Forks: 27
- Open Issues: 6
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# LanguageCrunch NLP Service docker image
Docker image: https://hub.docker.com/r/artpar/languagecrunch/
- **[Quickstart](#quickstart)**
- **[Endpoints](#endpoints)**
- **[Model Details](#model-details)**
- [Sentiment](#sentiment)
- [Entity extraction](#entity-extraction)
- [Sentence type detection](#sentence-type-detection)
- [Relation extraction](#relation-extraction)
- [Word look up](#word-look-up)
## Quickstart
Pull and run the Docker image, listening on port 8080:
```
$ docker run -it -p 8080:8080 artpar/languagecrunch
```
Example API call:
```bash
$ curl http://localhost:8080/nlp/parse?`echo -n "The new twitter is so weird. Seriously. Why is there a new twitter? What was wrong with the old one? Fix it now." | python -c "import urllib, sys; print(urllib.urlencode({'sentence': sys.stdin.read()}))"`
```
## Endpoints
### Sentence parse [Spacy]
`GET http://localhost:8080/nlp/parse?sentence=`
### Word lookup [Wordnet]
`GET http://localhost:8080/nlp/word?word=ask&pos=v`
### Coreference resolution [neuralcoref]
`GET http://localhost:8080/nlp/coref?sentence=`
## Model Details
### Sentiment
`sentence: The new twitter is so weird. Seriously. Why is there a new twitter? What was wrong with the old one? Fix it now.`
```json
{
"relations": [],
"sentences": [
{
"sentence": "The new twitter is so weird. ",
"sentence_type": "assertive",
"sentiment": {
"polarity": -0.18181818181818182,
"subjectivity": 0.7272727272727273
},
"root": {
"text": "is ",
"orth": 2
},
"pos": [
{
"text": "The new twitter",
"lemma": "the",
"pos": "DET",
"tag": "DT",
"dep": "nsubj",
.
.
.
```
### Entity extraction
- PERSON
- NORP
- FACILITY
- ORG
- GPE
- LOC
- PRODUCT
- EVENT
- WORK_OF_ART
- LAW
- LANGUAGE
- DATE
- TIME
- PERCENT
- MONEY
- QUANTITY
- ORDINAL
- CARDINAL
`Eg: Bill Gates, the founder of Microsoft, hosted a party last night`
```json
"entities": [
{
"text": "Bill Gates",
"label": "PERSON"
},
{
"text": "Microsoft",
"label": "ORG"
},
{
"text": "last night",
"label": "TIME"
}
]
}
```
### Sentence type detection
- assertive
- interrogative
- exclamatory
- negative
### Relation extraction
`Eg: Bill Gates, the founder of Microsoft, hosted a party last night`
```
"relations": [
{
"subject": "the founder",
"object": "Microsoft",
"relation": "ORG"
}
],
```
`Eg: Apple is looking at buying U.K. startup for $1 billion`
```[
{
subject: "N/A",
object: "U.K. startup",
relation: "GPE"
},
{
subject: "buying",
object: "$1 billion",
relation: "MONEY"
}
],
```
### Word look up
- Category of word
- Hypernyms - **colour** is a hypernym of **red**.
- Specific words of a category
- Holonyms - **red** is a holonym of **color**
- Synonyms to match
- Examples
- Word frames ( how the word is used )
- Coreference resolution
- Pronouns/references to nouns
`Eg: startle, verb`
```json
"results": [
{
"definition": "to stimulate to action",
"examples": [
"..startled him awake",
"galvanized into action"
],
"lemma_names": [
"startle",
"galvanize",
"galvanise"
],
"hypernyms": [
{
"definition": "surprise greatly; knock someone's socks off",
"examples": [
"I was floored when I heard that I was promoted"
],
"lemma_names": [
"shock",
"floor",
"ball_over",
"blow_out_of_the_water",
"take_aback"
]
}
],
"lemmas": [
{
"frame_strings": [
"Somebody startle somebody",
"Something startle somebody",
"Somebody startle somebody into V-ing something"
],
```
## Contributors
[List of contributors](contributors.md)