Multi-document summarisation is the process of producing a single summary of a collection of related documents. Much of the current work on multi-document text summarisation is concerned with the English language; relevant resources are numerous and readily available. These resources include human generated (gold-standard) and automatic summaries. Arabic multi-document summarisation is still in its infancy. One of the obstacles to progress is the limited availability of Arabic resources to support this research. When we started our research there were no publicly available Arabic multi-document gold-standard summaries, which are needed to automatically evaluate system generated summaries. The Document Understanding Conference (DUC) and Text Analysis Conference (TAC) at that time provided resources such as gold-standard extractive and abstractive summaries (both human and system generated) that were only available in English. Our aim was to push forward the state-of-the-art in Arabic multi-document summarisation. This required advancements in at least two areas. The first area was the creation of Arabic test collections. The second area was concerned with the actual summarisation process to find methods that improve the quality of Arabic summaries. To address both points we created single and multi-document Arabic test collections both automatically and manually using a commonly used English dataset and by having human participants. We developed extractive language dependent and language independent single and multi-document summarisers, both for Arabic and English. In our work we provided state-of-the-art approaches for Arabic multi-document summarisation. We succeeded in including Arabic in one of the leading summarisation conferences the Text Analysis Conference (TAC). Researchers on Arabic multi-document summarisation now have resources and tools that can be used to advance the research in this field.

Each individual has a distinguished writing style. But natural language generation systems pro- duce text with much less variety. Is it possible to produce more human-like text from natural language generation systems by mimicking the style of particular authors? We start by analysing the text of real authors. We collect a corpus of texts from a single genre (food recipes) with each text identified with its author, and summarise a variety of writing features in these texts. Each author's writing style is the combination of a set of features. Analysis of the writing features shows that not only does each individual author write differently but the differences are consistent over the whole of their corpus. Hence we conclude that authors do keep consistent style consisting of a variety of different features. When we discuss notions such as the style and meaning of texts, we are referring to the reac- tion that readers have to them. It is important, therefore, in the field of computational linguistics to experiment by showing texts to people and assessing their interpretation of the texts. In our research we move the thesis from simple discussion and statistical analysis of the properties of text and NLG systems, to perform experiments to verify the actual impact that lexical preference has on real readers. Through experiments that require participants to follow a recipe and prepare food, we conclude that it is possible to alter the lexicon of a recipe without altering the actions performed by the cook, hence that word choice is an aspect of style rather than semantics; and also that word choice is one of the writing features employed by readers in identifying the author of a text. Among all writing features, individual lexical preference is very important both for analysing and generating texts. So we choose individual lexical choice as our principal topic of research. Using a modified version of distributional similarity CDS) helps us to choose words used by in- dividual authors without the limitation of many other solutions such as a pre-built thesauri. We present an algorithm for analysis and rewriting, and assess the results. Based on the results we propose some further improvements.