Search Engine Query Clustering

Goal: Need to create a program clustering search engine queries (union queries per group) based on the issuance of a search engine (in this case, SERP Yandex). Also likely to need to consult (adjustment) TK. Input: There SERPs PS Yandex to the top 50 for the n queries (say for 500,000 queries). It should be on the basis of the issuance of the document address (url address) according to the request, to the clustering of search query data (the output should be a set of query). Dining clustering algorithms: 1) A simple algorithm. Putting it in the group who searches within the top 10 results in line 3 and a url regardless of the position taken by the data according to the request url. Queries for which there is no overlap of the count of a url with other requests count a separate file as a non-clustered. Image: [login to view URL] 2) The algorithms of the articles Search Engine Query Clustering using Top-k Search Results [login to view URL]~yhong/pub/[login to view URL], [login to view URL] jrwen / qc-www10.pdf. It is worth noting that in English is rather weak and I detail these algorithms do not understand. Data: Data is stored in a large text file. Data output can transform into another kind and provide artist. Given the number of requests and the complexity of the algorithm will have to think to write in C (at least the creators of a similar service [login to view URL] wrote it in C and ASM). Not exactly in PHP. Config computer for clustering(if necessary give access!) Intel Core i7-920 Quadcore, 48 gb RAM Output: Excel or csv file with groups of search engine queries. image: [login to view URL] Requirements for contractor: You at least have to be an experienced programmer. Be sure to sort out the complexity of the algorithms and evaluate how much time you will be able to work out the written program. Give preference to artists with the knowledge in the field of Information Retrieval, Data Mining. Sources necessarily allocated to me. Operation and Payment: Offering a 500 $. for the project. Me and my contractor are coordinating the prices and terms. If you and me are both satisfied with the conditions of it we will continue our work. In case of a serious disruption of pre-negotiated terms I reserve the right to refuse to pay for the work. Let us respect each other's time. I work with no prepayment! Partial prepayments are only possible in case of a clear demonstration of fulfillment of a certain part of the project (can be demonstrated or via TeamViewer). On completion of the project show me the project working on my data and the results of it. If everything is allright - I pay fully for your work, and then you turn it the project! This payment scheme is optimal for the contractor and for me.