In my previous post I talked about the work I carried out for my MSc research project. Here is the Wordpress plugin that implements a simplified version of the summarization system described there.

This plugin creates an automatic excerpt of a post when it is saved or published. The excerpt created is a coherent piece of text containing the most important sentences of the post.
After installing this plugin, every time you write a post and press the save or publish button an excpert is created for that post and stored in the database. This plugin is useful also if you don’t use excerpts in your blog: it’s always useful to have a summary of your post available just to remind yourself what the post is about (it works very well with old and long posts!).
Statistical natural language techniques are used to extract the most important sentences from a document. It’s straightforward to tweak the plugin to give more importance to certain features in your text.
The code of the plugin has been optimized for extensiblity in order to make it easy for contributor developers to implement a summarizer in their own language. At the moment only an English language summarizer is available, and an Italian one is on its way.
After activating the plugin here are simple steps that you might want to follow in order to have a nice excerpt to use in your blog:

1. Write your post
2. Press the save or publish button
3. An excerpt will be created. Now you can review it, modify it and save it as many times as needed
4. If you want the summarizer to create an excerpt again delete the existing one and press save or publish

Note that the plugin creates automatically an excerpt for a post only if there isn’t already an excerpt for that post.

Here is the abstract:

In recent years personalised summarisation has been a hot topic in Information Retrieval. An example of a personalisation technique is query-biased summarisation, where the importance of a sentence is estimated based on the amount of information about the query that it contains. An appealing implementation of this method is the one undertaken by many web search enginesthat provide snippets generated with sentences, or part of sentences, that contain query terms. However, the query string is not enough to represent the user’s interests and information need, and a more complex representation is needed. We present in this work a method to create a richer user model by analyzing the interaction of the user with a blogging system. This model is then exploited to create personalised summaries that contain information tailored to the user’s interests. The system also allows the user to extend initial summaries with broader summaries that contain novel information, and dynamically adapt to the model of the user’s information need. A system- centric and user-centric evaluation suggest that the proposed summarisation method performs better than query-biased summarisation in extracting relevant sentences from a document, thus showing that adaptive summarisation is an effective way to support the user in understanding the relevancy of a document to a particular topic.