emanjavacas/cosycat
Fork: 7 Star: 9 (更新于 1970-01-01 00:00:00)
license: EPL-1.0
Language: Clojure .
Collaborative Synchronized Corpus Annotation Tool
最后发布版本: 0.2.5-alpha ( 2018-04-17 22:19:15)
Cosycat (Collaborative Synchronized Corpus Annotation Tool)
For more documentation see the wiki.
Introduction
Cosycat is a web-based corpus query and annotation interface with a focus on multiuser synchronization and version control. The application is designed with an emphasis on modularity and reusability.
What can I do with Cosycat?
The major goal of Cosycat is to enable synchronized multiuser annotation of corpora. While Cosycat can absolutely be used for annotating typical NLP resources (such as Part-of-Speech tagging a newspaper corpus), it is mostly geared towards (preferrably team-based) Corpus Linguistic research aimed at statistically analyzing particular constructions that have to be first retrieved and annotated.
Cosycat allows multiple users to analyze the same corpus in an interactive way, keeping their annotations synchronized with each other and receiving live feedback from each other actions. Each change to the current body of corpus annotations (e.g. each time an annotation is introduced, modified or removed by a user) is recorded in a version control fashion which allows to easily backtrace the annotation history.
How does it work?
Cosycat is built taking advantage of modern browser features such as WebSockets or LocalStorage, recent developments in reactive UI programming (see the React framework), as well as the flexibility of NoSQL databases.
-
Architecture
Cosycat follows the following client-server architecture:
The client interacts with the local annotation database and with the (possibly) remote server via HTTP. This gives you the advantage of not having to host corpus yourself (although, this is certainly possible).
-
User interaction
User activity is routed through the server and then propagated to the rest endpoints (users) that are currently using the application. Cosycat informs users of each others activities using a built-in notification system.
-
Projects (access rights, pub/sub notifications)
A central organizatory piece of Cosycat is the "project". Projects serve multiple purposes:
- Control notification verbosity (subscribing/unsubscribing from projects)
- Grouping annotations into different
- Advanced: (delayed merge) use projects as local working copies.
-
Version Control
Currently, Cosycat allows you to keep track of all changes effected over the annotation database. This is done with a small module, that stores changes to a MongoDB document in a version-controlled collection using a traditional lock system for data consistency.
Citing
If you've found this work useful, please cite the following paper:
@inproceedings{Arevalo:2017:EAH:3078081.3078089,
author = {Ar{\'e}valo, Enrique Manjavacas and Petr{\'e}, Peter},
title = {Enabling Annotation of Historical Corpora in an Asynchronous Collaborative Environment},
booktitle = {Proceedings of the 2Nd International Conference on Digital Access to Textual Cultural Heritage},
series = {DATeCH2017},
year = {2017},
isbn = {978-1-4503-5265-9},
location = {G\öttingen, Germany},
pages = {9--14},
numpages = {6},
url = {http://doi.acm.org/10.1145/3078081.3078089},
doi = {10.1145/3078081.3078089},
acmid = {3078089},
publisher = {ACM},
address = {New York, NY, USA},
keywords = {Asynchronous Collaboration, Corpus Annotation, Groupware},
}
emanjavacas/cosycat同语言 Clojure最近更新仓库
2024-11-05 23:13:48 logseq/logseq
2024-11-04 23:45:19 penpot/penpot
2024-10-03 02:18:04 metabase/metabase
2023-02-20 20:09:25 adamtornhill/code-maat
2021-12-07 04:58:58 tonsky/FiraCode
1970-01-01 00:00:00 nickik/fscan