Konstantin Sokolov's repositories
lstm-from-scratch-java
LSTM-based encoder-decoder architecture written in Java from scratch
ccg-corpus
Tools and syntactic annotation workplace (Java, Swing) for OpenCorpora
CL-Toolbox
A Parsing-as-Deduction system that parses with different parsing algorithms and some related tools for educational purposes.
fca
Implementation of NextClosure algorithm (Ganter, 1987) for Formal Concept Analysis (FCA)
lm
Language modeling before it was mainstream
midterm
Sample assignment for my NLP course at Peter the Great St. Petersburg Polytechnical University
mp4ru
MaltParser for Russian
simple-lcg
A simplistic parser for Lambek Categorial Grammar based on term graphs, a variant of proof nets introduced in (Fowler 2016).
slides
Various TeX for my talks at Natural Language Understanding Reading Group (2013-2016)
tweets-geolocation
PyTorch implementation of UnicodeCNN architecture from (Izbicki et al. Geolocating Tweets... 2019) for short texts geolocation
hlmc
Hybrid Logic Model Checker (HLMC) and MCFull algorithm (Dragone, 2005) ported to Java
javafx-guice-mvp
Proof-of-concept architecture for JavaFX based on Dependency Injection, Model-View-Presenter (Passive View variant) and EventBus patterns
manifold-embeddings-s
A thin FastAPI wrapper over all-MiniLM-L6-v2 model for deployment in Docker. API similar to OpenAI Embeddings endpoint.
manifold-indexer-job
S3 (MinIO) indexer cronjob
manifold-search
Example client for vector search in Redis
open-synsem
Semantic interpretation of natural language commands with OpenCCG and Hybrid Logic
slides_2020
Slides for NLP Courses in 2020
TextClassification
A Text Classification API in Java originally developed by DigitalPebble Ltd. The API is independent from the ML implementations used and can be used as a front end to various ML algorithms. libSVM and liblinear are currently embedded.
TextRank-1
TextRank算法提取关键词的Java实现
topic-modeling-survey
Survey of papers on Topic Modeling and Automatic Topic Naming (in Russian)