A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
M Kagalkar, Ramesh
- Methodology of Context Centered Term Indexing Style Intended For Hindi Language Document Summarization
Authors
Source
Software Engineering, Vol 8, No 5 (2016), Pagination: 125-130Abstract
Document outline is beneficial since it will provide the summary of the initial document in an exceedingly shorter amount of your time. Document summarization largely uses the similarity between sentences within the document to extract the foremost salient sentences. The documents moreover because the sentences are indexed using ancient term indexing measures, that don't take the context into thought. Therefore, the sentence similarity values stay freelance of the context. The planned system consists of a context sensitive document compartmentalization model supported the Bernoulli model of randomness. The Bernoulli model of randomness has been accustomed notice the probability of the concurrences of 2 terms in passing big corpus. A replacement approach practice the lexical association between terms to relinquish a context sensitive weight to the document terms has been planned. The following compartmentalization weights are accustomed reason the sentence similarity matrix. The planned sentence similarity live has been used with the baseline graph-based ranking models for sentence extraction.
Keywords
Text Summarization, Lexical Association, Context Sensitive Indexing.
- A Methodology for Sign Language Video Analysis and Translation into Text in Hindi Language
Authors
Source
Fuzzy Systems, Vol 8, No 5 (2016), Pagination: 137-143Abstract
Sign language based communication is basically essential by consistently changing distinctive hand shapes and development by a signer. The proposed framework going to recognize the sign of gesture-based communication from a video stream of the signer and deciphers into relating words and sentences in Hindi. The proposed framework utilizes the image preparing strategies and synthetic intelligence procedures to get the target. To do the proposed task execution it utilizes picture handling strategies, for example, frame differencing based tracking, edge recognition, wavelet change, picture combination systems to area shapes in our videos. It moreover utilizes descriptors of elliptical Fourier for feature extraction of shape and also uses analysis for the purpose of feature set reduction as well as optimization. Extracted features are stored into the database and compared with given input video of the signer by a trained unclear inference system