In modern world, large data is available for every single topic and it’s really a hard process to know and take out the exactly relevant material when looking for it online and exactly this is where the text summarization is required. Text summarization is the process of filtering the most important information from the original source to reduce the length of the text document and automatic text summarization is the process of generating summaries of a document by using software and technological tools. There are two basic approaches to automatic text summarization which are:
Extraction-based summarization
In extraction-based summarization, a subset of most important words is extracted from the original text document and is combined to make a summary. It is like a highlighter that extracts the main information from a source text and highlights it. In extraction based summarization the extracted summary is composed of the highlighted or repetitive words, although the results may or may not be grammatically accurate always.
Abstraction-based summarization
In abstraction-based summarization, more advanced machine learning techniques are applied to shorten the length of original document and it creates a summary just like humans do. In abstraction based summarization novel sentences are created that may contain words which are not even a part of the original document. Scope for grammatical errors is almost nil in this type of summarization as it can generate new phrases and sentences and extracts the most important information from the source text.
To have more accurate and reliable summaries it is better to switch from extractive to abstractive summarization because contrary to extractive methods, abstractive techniques display summarized information in a coherent and reliable form that is both grammatically correct and easily readable.
Abstractive summarization method shows less stable results as compare to extractive summarization methods. But still it is believed that abstractive method approach is more promising in terms of generating human-like summaries. Therefore, more approaches are mushrooming in this field and offers new perspective from the computational, cognitive and linguistic points of view.
Though automatic text summarization is required in each and every field of work, but there are some common and important uses of it which includes:
- It reduces the reading time.
- It reduces searching time while researching as instead of reading the whole document, you can select by reading summaries.
- It reduces the searching time for business persons, analysts also as most of their time is spent just figuring out which document is relevant and which isn’t. By using summarizers, they can figure out the importance of a document before opening it.
- Along with business people, analysts, scholars and researcher, it also helps students and authors to generate abstracts of their research paper or book chapter.
