Abstract: A method and an apparatus for tracking changes in a collection of web documents, for example, provided by a web site. The web documents are retrieved at a first assigned point in time and a second assigned point in time. Then a similarity measure for a combination of a retrieved web document at a first assigned point in time and a retrieved web document at a second assigned point in time is calculated for determining pairs of corresponding web documents. By comparing said calculated similarity measure of a pair of corresponding web documents with predetermined thresholds for the similarity measure a change in the content of the corresponding web document between the first assigned point in time and second assigned point in time is detected. Instead of referring to identifiers like URLs for web pages the content similarities of web pages are considered. The proposed strategy facilitates the work of marketing analysts.

Abstract: A method for estimating a sentiment conveyed by the content of information sources towards an entity is presented. The sentiment is obtained with respect to a query context that may be specified, e. g. by specific terms or expressions, like a product or service name. A sentiment dictionary having a plurality of sentiment terms is provided, wherein each sentiment term has assigned a sentiment value, and at least one of said sentiment terms is associated to a group context. Text documents are screened for occurrences of sentiment terms that are associated to a group context corresponding to the query context. Calculating a sentiment score value is performed as a function of the occurrences of sentiment terms having a similar or same group context as the query context. The method may be carried out automatically without manual analysis of the actual semantic content of the text documents under consideration.

Abstract: An overall semantic similarity score value between pairs of named entities in a text corpus is obtained by calculating for at least one pair of named entities a plurality of corresponding pair similarity score values according to a first and at least a second classifier using electronic information sources. Each pair similarity score value of the pair of named entities per classifier is normalized by calculating a rank list per classifier, for example, for each named entity. The rank list holds each pair of named entities of the text corpus, wherein a rank of each pair of named entities within the rank list reflects the respective pair similarity score value. Further an arithmetic mean of the normalized pair similarity score value of each pair of named entities is calculated to provide the overall semantic similarity score value.

Abstract: A method and an apparatus for estimating a prestige of an entity, as for example a firm, company or name, is disclosed wherein a score value is assigned to an entity as a function of an occurrence of terms associated with said entity in search results. The search results are obtained by searching an information space such as the internet. This enables, for example, companies or divisions, to infer their public standing from an analysis of search results obtained through internet search engines. It is possible to compare a plurality of entities with respect to each other in an automated fashion.

Abstract: An overall semantic similarity score value between pairs of named entities in a text corpus is obtained by calculating for at least one pair of named entities a plurality of corresponding pair similarity score values according to a first and at least a second classifier using electronic information sources. Each pair similarity score value of the pair of named entities per classifier is normalized by calculating a rank list per classifier, for example, for each named entity. The rank list holds each pair of named entities of the text corpus, wherein a rank of each pair of named entities within the rank list reflects the respective pair similarity score value. Further an arithmetic mean of the normalized pair similarity score value of each pair of named entities is calculated to provide the overall semantic similarity score value.

Abstract: A method for estimating a sentiment conveyed by the content of information sources towards an entity is presented. The sentiment is obtained with respect to a query context that may be specified, e.g. by specific terms or expressions, like a product or service name. A sentiment dictionary having a plurality of sentiment terms is provided, wherein each sentiment term has assigned a sentiment value, and at least one of said sentiment terms is associated to a group context. Text documents are screened for occurrences of sentiment terms that are associated to a group context corresponding to the query context. Calculating a sentiment score value is performed as a function of the occurrences of sentiment terms having a similar or same group context as the query context. The method may be carried out automatically without manual analysis of the actual semantic content of the text documents under consideration.

Abstract: A method and an apparatus for tracking changes in a collection of web documents, for example, provided by a web site. The web documents are retrieved at a first assigned point in time and a second assigned point in time. Then a similarity measure for a combination of a retrieved web document at a first assigned point in time and a retrieved web document at a second assigned point in time is calculated for determining pairs of corresponding web documents. By comparing said calculated similarity measure of a pair of corresponding web documents with predetermined thresholds for the similarity measure a change in the content of the corresponding web document between the first assigned point in time and second assigned point in time is detected. Instead of referring to identifiers like URLs for web pages the content similarities of web pages are considered. The proposed strategy facilitates the work of marketing analysts.

Abstract: A method and apparatus for comparing entities, such as companies or trademarks, based on available pricing and feature information regarding goods or services provided by the companies. Web services, e.g. price finding robots, are employed to obtain pricing and feature information for similar products from different companies. Products having similar features but derive from different companies can be grouped to clusters and be analyzed with respect to average pricing levels. Data formats for pricing and feature information stemming from different web services may be automatically retrieved and mapped into a common format. The method allows to evaluate systematic deviations in the company's pricing policies or to detect if companies do not have a matching product portfolio. One can also estimate the prestige or acceptance of a company or brand as a function of prices tolerated by the underlying market. This may facilitate marketing strategies.

Abstract: A method and an apparatus for estimating a prestige of an entity, as for example a firm, company or name, is disclosed wherein a score value is assigned to an entity as a function of an occurrence of terms associated with said entity in search results. The search results are obtained by searching an information space such as the internet. This enables, for example, companies or divisions, to infer their public standing from an analysis of search results obtained through internet search engines. It is possible to compare a plurality of entities with respect to each other in an automated fashion.

Abstract: Textual content is extracted from hypertext documents by generating for each text document a pruned document model tree of merged text nodes by removing selected tag nodes from a document model tree of the text document, calculating for each merged text node of the pruned document model tree a set of text features which are compared with predetermined feature criteria to decide whether the merged text node is an informative merged text node, and assembling the informative merged text nodes to generate a text file containing the textual content.

Abstract: The invention relates to a method and an apparatus for performing a drill-down operation on a text corpus comprising documents, using language models for key phrase weighting, said method comprising the steps of weighting key phrases occurring both in a foreground language model, which contains a selected document cluster of said text corpus, and in a background language model, which does not contain said selected document cluster, by calculating for each key phrase a key phrase weight comprising a ratio between the foreground weight of said key phrase and a background weight of said key phrase, and assigning documents of the foreground language model to cluster labels which are formed by key phrases having high calculated key phrase weights.