2 ;;Quote: indexing core for constructing a document-level index with ranked query evaluation; refinements for reorganization, phrases, and compression; includes bibliography

8 ;;Quote: use inverted files for text query evaluation; better than suffix arrays and signature files

8 ;;Quote: all terms should be indexed; including numbers, URL tokens, and stopwords

13 ;;Quote: build two indices, a word-level index for phrase and Boolean searches and a document-level index for ranked queries

13 ;;Quote: for phrase queries, use an index for word pairs that begin with a common word; the three commonest words halves phrase querying time

16 ;;Quote: merge-based index construction scales well; 100MB of memory, can minimize disk space overhead, only one parsing pass, compression effective

32 ;;Quote: while fast for grep, suffix arrays do not scale well; no compression, 1.7x memory-resident data, no ranked queries

32 ;;Quote: grep pattern matching by suffix array of vocabulary or by inverted index of digrams or trigrams

33 ;;Quote: problems with signature files -- false matches are linear in the collection size, large index, more disk accesses for short queries, and no ranked queries

35 ;;Quote: index the anchor text as a title of the target page; helps identify a document; good for Web searching